Arbuscular mycorrhizal colonization does not improve root hydraulic supply in tomato and pea

Abstract

Arbuscular mycorrhizal (AM) fungi are known to enhance plant drought tolerance, but the physiological mechanism behind this benefit remains unclear. One explanation is that AM colonization improves root hydraulic conductance (K_r), thereby facilitating more efficient water uptake under soil drying, though this mechanism remains highly debated. Here, we measured K_r in tomato (Solanum lycopersicum L.) and pea (Pisum sativum L.) with and without AM using a noninvasive rehydration technique under soil drying, and this was complemented with the evaporative flux method under hydrated conditions. AM colonization was manipulated either through soil sterilization or by using nonmycorrhizal mutants, ensuring precise control of AM status. In both species, AM colonization had no positive impact on K_r under both well-hydrated and drought conditions. The finding suggests that the improved drought performance often observed in AM-colonized plants is not due to enhanced root water transport capacity. Instead, AM-induced benefits under drought may be mediated by other physiological adjustments.

Arbuscular mycorrhizal symbiosis had no effect on root hydraulic conductance in 2 major crops, suggesting it does not improve drought tolerance by increasing water uptake.

Introduction

Roots play a major role in supplying the shoot canopy with water, maintaining productivity and survival. Root water uptake capacity has been reported to be highly sensitive to the initial stages of soil drying, reducing plant water uptake and gas exchange (Rodriguez-Dominguez and Brodribb 2020; Bourbia et al. 2021). Root hydraulic conductance (K_r) reflects the efficiency of plant roots to conduct water from the root–soil interface to the root collar (Steudle 2000). The dynamics of K_r during soil drying have been shown to be influenced by several factors such as root structure, root length, and aquaporin activity (Vadez 2014; Jeanguenin et al. 2016). However, the role of rhizosphere microorganisms on K_r dynamics under progressive soil drying is still poorly understood.

Rhizosphere microorganisms play an important role in plant growth, health, and ecosystem sustainability. This includes beneficial microbes that promote plant nutrient uptake, as well as detrimental microbes that challenge plant health (Mendes et al. 2013; Omotayo and Babalola 2021). Arbuscular mycorrhizal fungi (AMF) belong to the phylum Glomeromycota and form symbioses with 71% of all land plants (Brundrett and Tedersoo 2018). Arbuscular mycorrhizal (AM) symbioses promote plant nutrient uptake (Wahab et al. 2023), plant growth, and, in many cases, drought tolerance (Yan et al. 2022; Obase et al. 2023). While the role of AM associations in the uptake of nutrients is well understood, the ability of AM-colonized plants to enhance water uptake under soil drying conditions is still a matter of debate (reviewed by Abdalla et al. 2023).

In addition to root hydraulics, plants employ a range of adaptive strategies to resist drought stress. These include hormonal regulation, such as abscisic acid (ABA)-mediated stomatal closure and the modulation of other hormone pathways; osmotic adjustment through the accumulation of osmolytes like proline, sugars; and morphological modifications such as increased root length (Fang and Xiong 2015; Lim et al. 2015; Kaur et al. 2016; Ruiz-Lozano et al. 2016; Naikwade 2023). These mechanisms help to preserve cellular hydration and maintain productivity under water stress. AM symbiosis has been proposed to influence some of these processes, for example, by modulating ABA signaling or improving soluble sugar contents that support osmotic adjustment, thereby contributing to plant drought tolerance through multiple mechanisms (Xu et al. 2018; Yang et al. 2025).

Plants with AM symbiosis have been reported to exhibit higher photosynthesis, stomatal conductance (g_sc), and transpiration compared to corresponding nonmycorrhizal plants under both well hydrated and drought conditions (Kuyper et al. 2021). For example, a meta-analysis conducted by Augé et al. (2016), integrating findings from 1,019 studies, reported that AM-colonized plants exhibited a 49% increase in carbon exchange rate, along with a 28% and 26% rise in g_sc and transpiration, respectively, compared to non-AM plants. Similarly, a study conducted by Kakouridis et al. (2022) showed that oats with AMF transpired twice as much as those without, indicating that AMF may extend the roots access to water. The higher g_sc of AM plants during soil drying suggested that AMF may access water unavailable to non-AM plants, potentially by the hyphal networks taking up water (Kakouridis et al. 2022). While studies have highlighted the positive effects of AMF on aboveground physiological function, less attention has been paid to their influence on the function of root hydraulics under drought conditions.

One possible explanation for the enhanced drought tolerance of AM-colonized plants is an increase in K_r. Ideally, this hypothesis should be tested using a direct, noninvasive method to measure K_r continuously from well-hydrated to drought conditions. While previous studies have examined the impact of AM colonization on root hydraulics, their results have varied with some reporting increased K_r, others showing no effects on K_r (Allen 1982; Levy et al. 1983; Graham and Syvertsen 1984; Gonzalez-Dugo 2010; Abdalla and Ahmed 2021; Pauwels et al. 2023). The discrepancy remains elusive and may stem from differences in species-specific behaviors or experimental design and measurement approaches used to examine K_r. For example, in many cases K_r is measured indirectly using the evaporative flux method, which assumes steady-state plant water potential and transpiration (Gonzalez-Dugo 2010; Abdalla and Ahmed 2021). The steady-state assumption may not hold during progressive soil drying, as plant and soil water potentials as well as transpiration continuously decline throughout the day (Bourbia et al. 2022; Bourbia et al. 2025). Other techniques involve invasive approaches, such as the root pressure probe and the pressure chamber (Knipfer and Steudle 2008 ), provide valuable insights but involve steps that may alter root functions or introduce errors due to root damage and do not reflect natural conditions (Li and Liu 2010). Furthermore, these studies focused on well-hydrated conditions or compared only extreme wet and dry states (reviewed by Kakouridis et al. 2022), without examining a progressive decline in soil moisture where plants actively regulate water use and carbon gain.

To address these limitations and provide robust evidence for the influence of AM symbiosis on root water transport under realistic drought scenarios, we use a noninvasive rehydration method for directly measuring K_r based on the relaxation kinetics of stem water potential during hydration of plants from different levels of soil water stress (Rodriguez-Dominguez and Brodribb 2020; Bourbia et al. 2021). By applying this noninvasive rehydration method, K_r dynamics in AM and non-AM plants were compared. To clarify the effects of AM, we combined 2 complementary strategies to control AMF: inoculation with active or sterilized AMF inoculum, and the use of nonmycorrhizal mutants. The study was conducted with 2 herbaceous species, Solanum lycopersicum L. (tomato) and Pisum sativum L. (pea), under a large range of soil water stress, from well-hydrated conditions to moderate stress that was sufficient to cause a substantial decline in K_r. This combination of a noninvasive rehydration method, 2 AMF control strategies and multiple cultivars of tomato and pea provides an ideal framework for understanding how AM symbiosis influences root hydraulics during dehydration and recovery.

Results

AM colonization has no impact on K_r in wild-type tomato under hydrated or drought conditions

The relationship between K_r normalized by projected leaf area (Fig. 1a) or root fresh weight (Fig. 1b) and Ψ_stem under different water stress conditions was observed in wild-type tomato (cultivar M82) colonized with arbuscular mycorrhiza (+AM) and without arbuscular mycorrhiza (−AM). Successful AM colonization occurred in +AM inoculated tomato plants and was zero in −AM plants (Fig. 1c). There was no significant difference between wild-type tomato plants +AM and −AM for leaf area, shoot and root weight, or the ratio of shoot-to-root (Supplementary Figure S1).

Figure 1.

Root hydraulic conductance (K_r) in Solanum lycopersicum L. wild type (M82) with arbuscular mycorrhiza (+AM) and without arbuscular mycorrhiza (−AM). (a and b) Scatter plot showing K_r normalized by (a) projected leaf area and (b) normalized by root fresh weight under hydrated and drought conditions (−Ψ_stem = 0–1.5 MPa). The shaded regions represent confidence intervals (95%) for each treatment group, with the fitted lines indicating the trend of K_r under drought conditions in relation to −Ψ_stem. Large error bars at −Ψ_stem = 0 indicate mean ± standard error under hydrated conditions; solid and dashed lines represent wild-type +AM and wild-type −AM treatments, respectively; (c) percentage of root colonized by AM. The center line represents the median; box limits indicate the upper and lower quartiles; whiskers extend to 1.5× the interquartile range; points represent individual biological replicates, with points beyond the whiskers considered outliers. For (a–c), data points represent individual replicate samples. Asterisks indicate significant differences between treatments (t-test, **P < 0.01). The sample size of M82 ± AM under drought or hydrated conditions is n = 14 to 15.

K _r decreased linearly in tomato with declining Ψ_stem in all treatments, whether K_r was normalized by projected leaf area or root fresh weight (Fig. 1a, b). Under hydrated conditions (Ψ_stem = 0 MPa), there was no significant difference in K_r in wild-type tomato plants colonized with AM fungi or not colonized with AM fungi (t-test: P-value = 0.571/0.839 for K_r under hydrated conditions normalized by projected leaf area or root fresh weight; Fig. 1a, b). Similarly, across a wide range of Ψ_stem, the presence or absence of AM colonization had no significant impact on K_r dynamics during soil drying (Fig. 1a, b; statistical analysis using the MuMIn package in R, with model selection result from Equation (3): ΔAICc = 0.00/0.00 in Model 1 and 3, respectively, Supplementary Table S1).

K _r is similar in AM-colonized wild-type tomato and nonmycorrhizal mutant rmc under hydrated and drought conditions

To validate the findings from the experiment comparing +AM and −AM in wild-type tomato cultivar M82 using the sterilization method (Fig. 1), the wild-type tomato cultivar 76R and its nonmycorrhizal mutant rmc were tested. As expected, rmc could not establish AM symbiosis and displayed no AM colonization, while wild-type 76R was well colonized (Fig. 2c). There was no significant difference between wild-type 76R and rmc tomato plants in leaf area, shoot and root weight, or the ratio of shoot-to-root (Supplementary Figure S1). The rmc mutant plants exhibited similar K_r dynamics in response the decline in Ψ_stem from −0 to −1.5 MPa as observed in wild-type plants (Fig. 2a, b; ΔAICc = 0.00/0.00 in Model 5 and 7, respectively, Supplementary Table S1). Under well-watered conditions (Ψ_stem = 0 MPa), no significant differences in K_r were observed between the 2 genotypes (t-test: P-value = 0.862/0.111 for K_r under hydrated conditions normalized by projected leaf area or root fresh weight; Fig. 2a, b).

Figure 2.

Root hydraulic conductance (K_r) in Solanum lycopersicum L. wild type (76R) colonized by mycorrhizal fungi (+AM) and respective nonmycorrhizal mutant rmc (−AM). (a and b) Scatter plot showing K_r normalized by (a) projected leaf area and (b) normalized by root fresh weight under hydrated and drought conditions (−Ψ_stem = 0–1.5 MPa). The shaded regions represent the confidence intervals (95%) for each treatment group, with the fitted lines indicating the trend of K_r under drought conditions in relation to −Ψ_stem. Large error bars at −Ψ_stem = 0 indicate mean ± standard error under hydrated conditions; solid and dashed lines represent wild-type +AM and mutant rmc −AM, respectively; (c) percentage of root colonized by AM. The center line represents the median; box limits indicate the upper and lower quartiles; whiskers extend to 1.5× the interquartile range; points represent individual biological replicates, with points beyond the whiskers considered outliers. For (a–c), data points represent individual replicate samples. Asterisks indicate significant differences between treatments (t-test, **P < 0.01). The sample size of 76R and rmc under drought or hydrated conditions is n = 9–19.

K _r is similar in AM-colonized wild-type pea and nonmycorrhizal mutant sym19 under hydrated and drought conditions

To investigate the impact of AM symbiosis on K_r in pea, wild-type cultivar Frisson was compared with its nonmycorrhizal mutant sym19 under both hydrated and drought conditions. Due to the mutation, sym19 is unable to establish AM symbiosis and displays zero AM colonization (Fig. 3c). In pea, AM symbiosis did not have a positive impact on K_r. Under hydrated conditions, K_r was not significantly different between wild-type pea and nonmycorrhizal sym19 mutant plants (t-test: P-value = 0.206/0.494 for K_r under hydrated conditions normalized by projected leaf area or root fresh weight; Fig. 3a, b). Interestingly, under drought conditions, a small but significant increase in K_r was observed in the nonmycorrhizal mutant sym19 compared to the AM-colonized wild-type pea cultivar (Fig. 3a, b). This was the case if K_r was normalized by projected leaf area or root fresh weight (ΔAICc = 0.00/0.27/0.00/0.90 in Model 9, 10, 12, and 13, respectively, Supplementary Table S1). It is important to note that the shoot fresh weight of sym19 was significantly lower than wild type (P < 0.01), despite similar root sizes (Supplementary Figure S1). Thus, it is possible that the small increase in K_r in the nonmycorrhizal mutant sym19 might be caused by smaller shoot size compared to wild-type pea.

Figure 3.

Root hydraulic conductance (K_r) in Pisum sativum L. wild-type (Frisson) and respective nonmycorrhizal mutant sym19 (−AM). (a and b) Scatter plot showing K_r normalized by (a) projected leaf area and (b) normalized by root fresh weight under hydrated and drought conditions (−Ψ_stem = 0–2.5 MPa). The shaded regions represent the confidence intervals (95%) for each treatment group, with the fitted lines indicating the trend of K_r under drought conditions in relation to −Ψ_stem. Large error bars at −Ψ_stem = 0 indicate mean ± standard error under hydrated conditions; solid and dashed lines represent wild-type +AM and mutant sym19 −AM, respectively; (c) percentage of root colonized by AM. The center line represents the median; box limits indicate the upper and lower quartiles; whiskers extend to 1.5× the interquartile range; points represent individual biological replicates, with points beyond the whiskers considered outliers. For (a–c), data points represent individual replicate samples. Asterisks indicate significant differences between treatments (t-test, **P < 0.01). The sample size of Frisson and sym19 under drought or hydrated conditions is n = 11–13.

The relationship between AM root colonization and K_r

To determine if there was any relationship between the extent of the root colonized by AM fungi and K_r, the percentage of root colonization by arbuscules was compared to K_r normalized by projected leaf area across different genotypes in tomato and pea (Fig. 4). No correlation between the percentage of the root colonized by arbuscules and K_r normalized by projected leaf area was found across both wild-type lines of tomato (M82 and 76R). In the tomato wild-type M82, no significant correlation was found between the percentage of arbuscule colonization and K_r, as indicated by the R² value close to zero (R² = 1e−06; Fig. 4a). Both M82 plants colonized by AM (+AM) and not colonized due to soil sterilization (−AM) displayed a wide distribution of K_r, with no obvious trend that could be explained by the level of arbuscule colonization. In the wild-type tomato 76R and respective nonmycorrhizal mutant rmc, no rmc plants were colonized and at least some individual wild-type 76R plants displayed low levels of colonization, but there was no evidence of a link between the percentage of the root colonized by arbuscules and normalized K_r, with a low R² value of 0.088582 (Fig. 4b). Similarly, no correlation between arbuscule colonization and normalized K_r was found in the wild-type pea Frisson and respective nonmycorrhizal mutant sym19, with an R² value of 0.078559 (Fig. 4c).

Figure 4.

Relationship between the percentage of the root colonized by arbuscules and root hydraulic conductance (K_r) normalized by projected leaf area across 3 genotypes in Solanum lycopersicum L. and Pisum sativum L. (a) Tomato M82 (wild type), with arbuscular mycorrhiza (+AM) and without arbuscular mycorrhiza (−AM). (b) Tomato 76R (wild type) and its nonmycorrhizal mutant, rmc. (c) Pea Frisson (wild type) and its nonmycorrhizal mutant, sym19. R² values are displayed in the top-right corner to indicate the goodness of fit for each regression model.

There was also no relationship found between K_r normalized by projected leaf area and total root colonization (the percentage of root colonized by any fungal structure; arbuscules, hyphae, and vesicles) across tomato or pea (Supplementary Figure S2). Similarly, no relationship was found between K_r normalized by root fresh weight and the percentage of root colonization by arbuscules (Supplementary Figure S3) or by total colonization (Supplementary Figure S4).

AM symbiosis does not influence C_plant in pea or tomato

To investigate whether AM symbiosis influences whole-plant capacitance, C_plant, we compared C_plant between plants with and without AM symbiosis across different species and genotypes. We hypothesized that AM symbiosis might enhance the plant's capability to hold water, causing an increase in C_plant. However, no significant differences were observed in C_plant between +AM and −AM for any genotype (Supplementary Figure S5), suggesting that the presence of AM symbiosis did not substantially influence C_plant in pea or tomato. For wild-type tomato cultivar M82, there was no significant change in C_plant when the +AM and −AM conditions were compared (Supplementary Figure S5). For nonmycorrhizal mutants, rmc and sym19, which could not form AM symbiosis even in the presence of AMF, no differences in C_plant were found in these mutants and respective wild types, 76R and Frisson (Supplementary Figure S5). These consistent results across species suggested that AM symbiosis did not contribute to C_plant, whether normalized by leaf area or root fresh weight, under the conditions tested.

Discussion

AMF are widely recognized for their ability to enhance plant drought tolerance (Ruiz-Lozano et al. 1995; Boomsma and Vyn 2008; Das and Sarkar 2024). However, the direct evidence linking AM colonization to increased K_r remains ambiguous. In previous studies, AM effects on K_r were typically examined under either wet or severely dry conditions, with little focus on the intermediate range where plants actively regulate water uptake. In this study, we integrated genetic and sterilization approaches across multiple cultivars of tomato and pea, to assess the influence of AM on K_r using a noninvasive rehydration method across a wide range of soil moisture conditions. By using a highly controlled experimental design, our study provides consistent and definitive evidence that AM did not significantly affect K_r in either tomato or pea under either hydrated and drought conditions. Thus, the commonly observed drought tolerance in AM plants is likely due to indirect physiological changes facilitated by AM rather than increased root water uptake efficiency.

It is worth pointing out that K_r can be influenced by traits such as root mass, root surface area, and root-to-shoot ratio, which can pose a challenge when comparing AM and non-AM plants, as non-AM plants are frequently smaller due to limited nutrient uptake. However, in our tomato studies, no differences were observed in leaf area, root mass, or root-to-shoot ratio between AMF and non-AM plants, allowing for a direct comparison of K_r between the 2 groups. Interestingly, a higher K_r normalized by projected leaf area and root fresh weight was observed in nonmycorrhizal pea mutant sym19 compared to wild-type pea cultivar Frisson under drought conditions (Fig. 3a, b, Supplementary Table S1: Model 9, 10, 12, and 13). This suggested AM symbiosis not only failed to enhance K_r under drought stress but also resulted in slightly lower K_r compared to the non-AM mutant sym19. The physiological data (Supplementary Figure S1a, b) showed significant reduction in projected leaf area and shoot fresh weight in sym19 compared to wild-type Frisson, while the root fresh weight and ratio of shoot and root remained unchanged (Supplementary Figure S1c, d). Nonmycorrhizal mutant sym19, compared to the wild-type Frisson, failed to establish symbiosis with AMF (Albrecht et al. 1999), possibly leading to reduced nutrient uptake. The smaller size of aboveground tissue may have reduced the water demand in sym19, indirectly contributing to its higher K_r. Indeed, other studies have found that AM colonization influences shoot biomass (Zhang et al. 2011; Qin et al. 2022). This suggests that K_r in pea was influenced by aboveground traits, rather than changes in the root system or AMF colonization. Importantly, AM symbiosis clearly did not improve K_r in pea.

AM colonization has been suggested to modify soil water retention properties and soil–root hydraulic interactions, potentially through altering the hydraulic conductance of the soil by binding the soil to the root surface, reducing the loss of hydraulic conductivity at the root–soil interface (Miller and Jastrow 2000; Bitterlich et al. 2018a). However, our results clearly demonstrate that any such influence does not extend significantly to internal root hydraulic conductance (the conductance from the root surface to the xylem). Instead, AM-mediated changes in soil properties may help maintain water flow from the bulk soil to the root surface, thereby sustaining some level of transpiration during drying, even when K_r is low in AM plants. This could explain the higher transpiration rates and drought tolerance often reported in AM plants under drying conditions (Bitterlich et al. 2018b).

The expression of aquaporins, which facilitate water movement across cell membranes and play a key role in regulating root water transport capacity, especially under water-limited conditions (Prado and Maurel 2013; Maurel et al. 2015), has become recognized as a major determinant of K_r. If AMF colonization upregulates the activity of aquaporins, an increase in K_r would be expected in colonized plants. However, the fact that K_r was not increased by AM colonization in our experiments suggests that aquaporin-mediated water transport was not significantly upregulated in pea or tomato AM plants. In fact, under drought, AM plants may downregulate aquaporins to reduce water loss (Zou et al. 2019). In addition to aquaporins, we can rule out AM-induced changes in root anatomical traits such as suberization of the endodermis, xylem development, and radial resistance as they would also influence K_r (Steudle 2000). Increased suberization under drought can reduce apoplastic flow and thereby limit K_r (Enstone et al. 2002; Franke and Schreiber 2007; McLean et al. 2011), regardless of AM colonization.

Labeled water uptake has been used to understand the contribution of AM hyphae to root water supply. Kakouridis et al. (2022) used a 2-chamber system in oats to track water transport by fungal hyphae, using ¹⁸O-labeled water as well as a fluorescent dye. They found that fungal hyphae directly transported 34.6% of transpired water via the extracellular pathway. In contrast, Püschel et al. (2020) used the same methodological approach in Medicago truncatula and found that water transported by AMF hyphae was relatively small compared to the plant's overall water demands. Although these and other studies (eg Khalvati et al. 2005 ) have shown the potential for AMF hyphae to take up and transport water, the small diameter of the hyphae imposes biophysical limitations on volume flow, making their direct contribution to K_r likely negligible under well-hydrated conditions and during soil drying (Rengel and Djalovic 2021). Our results clearly indicate that any direct contribution of AMF hyphae to internal K_r is negligible.

Although our results indicate that AMF did not significantly alter K_r in either tomato or pea, other studies have reported contrasting outcomes linked to AMF genotype and stress severity (Chitarra et al. 2016; Ruiz-Lozano et al. 2016). Such variability makes cross-study comparisons difficult. In our study, however, the same AMF strains and plant varieties were used within each species, thereby minimizing variation due to plant and fungal genetic background. Regarding drought intensity, the stress treatments applied covered a range of water potentials, which allowed compare responses across moderate to severe stress levels.

Our study suggests the enhanced drought tolerance frequently reported in AM-colonized plants (Zhang et al. 2019; Huang et al. 2020; Yang et al. 2025) is not due to increased K_r. Alternative explanations for AM-induced drought tolerance should be considered. Among the possible mechanisms, hormonal regulation and improved nutrient acquisition are likely key factors. AMF-induced modulation of hormones has been associated with improved performance of AM plants under drought compared to non-AM plants (reviewed by Cheng et al. 2021), such as strigolactones and ABA (Ruiz-Lozano et al. 2016). In addition, improved nutrient acquisition in AM plants, particularly phosphorus, could facilitate osmotic adjustment and metabolic balance during drought compared to non-AM plants (Püschel et al. 2021; Ullah et al. 2022). Supporting this view, some transcriptomic studies have shown that AM symbiosis modulates drought-responsive genes in host roots, including those involved in hormone signaling, osmotic adjustment, and secondary metabolite pathways (Balestrini et al. 2019; Xie et al. 2023, reviewed by Wang et al. 2023). This highlights the multiple mechanisms by which AM symbiosis can contribute to plant drought tolerance.

Conclusion

Using a robust experimental design, our study shows that AMF does not enhance K_r under either hydrated or drought conditions in tomato and pea. This is significant because it challenges the assumption that AMF improves drought tolerance primarily through enhanced K_r. Instead, our data indicate that in the distantly related species tomato and pea that AM did not influence K_r and it is likely that this is likely to also occur in other angiosperms. This suggests that the improved drought tolerance observed in AM-colonized plants is not driven by changes in K_r. AMF-induced drought tolerance likely depends on a range of mechanisms, including hormonal regulation and nutrient uptake.

Materials and methods

Plant material and growth conditions

Tomato (S. lycopersicum L.) and pea (P. sativum L.) were used in the experiments. Two wild-type genotypes of tomato M82 and 76R cultivar were used, along with the corresponding reduced AM colonization mutant (rmc) on 76R background. M82 and 76R formed AM symbioses in the presence of AM fungi, while rmc is the nonmycorrhizal mutant that has a large deletion that contains the CYCLOPS/IPD3 gene (Prihatna et al. 2018). The CYCLOPS/IPD3 protein plays a crucial role in the symbiotic signaling pathway, regulating the transcription of genes essential for establishing and functioning of AM symbioses (Yano et al. 2008). Tomato seeds were germinated in potting mix in punnet pots and transplanted after 3 wk in 0.5-liter (L) pots containing a substrate mixture of gravel, vermiculite, and live mycorrhizal inoculum; with a ratio of 2:2:1 or 2:2:2. The mycorrhizal inoculum used in this study came from the living corn (Zea mays) pot culture inoculated with the spores of Rhizophagus irregularis (INOQ Advantage, INOQ GMBH, Germany). Corn was planted in the substrate with 2-L pots (gravel:vermiculite:live mycorrhizal inoculation = 2:2:1) grown under natural light with an 18-h photoperiod. Corn cultures were assessed before using and consistently exhibited root colonization rates greater than 50% of root length colonized. 76R and rmc mutants were planted into nonsterilized AM pots (+AM). M82 plants were planted into nonsterilized AM pots (+AM) or into AM pots that were sterilized (−AM) by autoclaving at 121 °C for 75 min. To reintroduce any other microbes presented in +AM pots, a soil filtrate was created by mixing 2 +AM pots with 1 L of Milli-Q water. The water layer was carefully tipped off from the mixture after sedimentation. The soil filtrate was passed through a micro-cloth with a pore size of 20 μm using a vacuum to remove any AM fungal spores from the filtrate and 20 mL of this filtrate was applied to −AM pots at the time of transplantation. Tomato plants were supplied with a modified Long Ashton nutrient solution (Hewitt 1966) with 3.7 mM KNO₃ and 0.15 mM NaH₂PO₄ twice a week after transplanting, at a volume of 40 mL per pot per application. Tomato was used for hydraulic experiments 4 to 5 wk after transplanting.

Pea studies were carried out with wild-type Frisson and corresponding nonmycorrhizal sym19 mutant. Frisson forms AM symbioses in the presence of AM fungi and the sym19 line is mutated in SYMRK/DMI2 gene, which encodes a leucine-rich repeat receptor-like kinase essential for activating the signaling pathway required to establish nodulation and AM symbiosis (Stracke et al. 2002). Thus, sym19 is unable to form nodules or AM symbiosis even in the presence of rhizobia or AMF (Geurts and Bisseling 2002; Bersoult et al. 2005). Pea seeds were grown in 0.5-L pots with potting mix (potting mix:live mycorrhizal inoculum as outlined above, 4:1). Pea was provided with a modified Long Ashton nutrient solution (Hewitt 1966) with 4.8 mM KNO₃, 2.6 mM Ca(NO₃)₂·4H₂O, and 0.05 mM NaH₂PO₄ once a week, with 40 mL of solution applied per pot each time. All the plants were grown in high-intensity light cabinet (c. 200 lm m⁻²) at 25/22 °C day/night with an 18-h photoperiod. Pea plants were used for hydraulic experiments 3 to 4 wk after planting.

K _r in pea and tomato under well-hydrated conditions

K _r in well-watered plants (with soils at field capacity) was measured using the evaporative flux method under steady-state conditions in well-watered, unstressed plants as described in Equation (1) (Tsuda and Tyree 2000):

$$K_{\mathrm{r}}=E/({\mathrm{\Psi }}_{\mathrm{soil}}-{\mathrm{\Psi }}_{\mathrm{stem}})$$ (1)

E is average canopy transpiration, which was measured gravimetrically in each plant under hydrated conditions with a balance (model: VIBRA ALE-6202, 6,200 G × 0.01 G) at 30-min intervals at midday between 12:00 and 14:00 (during which time transpiration and Ψ_stem were at steady state in both species). This was supported by graph generated from the optical dendrometer measurements of an example plant, which showed that tissue shrinkage remained stable during midday, indicating that Ψ_stem was at steady state (Supplementary Figure S6). E was normalized by the projected whole-plant leaf area measured at the end of the experiment. Ψ_stem was measured with a pressure chamber immediately following E measurements. These measurements were taken on nontranspiring leaves that had been covered with wet paper towel and aluminium foil for at least 2 h prior to E measurements to ensure equilibration between leaf and stem water potential. Ψ_soil, the soil water potential, was assumed to be close to 0 MPa because soils were watered to field capacity before measurements.

K _r in pea and tomato in response to water stress

The response of K_r (mmol s⁻¹ m⁻² MPa⁻¹) to increasing soil drying was measured under nonsteady-state conditions by examining the kinetics of Ψ_stem relaxation during instantaneous rewetting of the soil, thereby overcoming any soil hydraulic resistance such that only K_r remained as a limiter of rehydration kinetics. Rehydration experiments were conducted on intact plants dehydrated to different levels of water stress. Biological replicates per species and treatment ranged from 12 to 19. Water deficits were created by withholding watering for different periods of time while continuously monitoring Ψ_stem using optical dendrometers. Optical dendrometers were attached to petioles for tomato and tendrils for pea to continuously monitor changes in tissue width and were used as a proxy for Ψ_stem dynamics. A linear relationship between Ψ_stem and width of tomato petiole/pea tendril was consistent in all plants (Supplementary Figure S7). For each plant, the dendrometer-derived width of tomato petiole/pea tendril was calibrated against Ψ_stem measured periodically on covered leaves with a Scholander pressure chamber (PMS Instruments, Albany, OR, USA) once the soil began to dry. After reaching the desired water stress level and before rewatering, plants were placed in the dark at high humidity (>80%) for at least 2 to 3 h to ensure stomatal closure and allow Ψ_stem and Ψ_soil to come to equilibrium (as observed with the dendrometer). Therefore, Ψ_soil was assumed to be equivalent to Ψ_stem before rewatering (after equilibration in the dark). Then the plants were rewatered and kept under the same conditions until Ψ_stem recovered to maximum levels (close to 0) (Supplementary Figure S6). K_r was then determined by assuming the rehydrating plant behaved as a capacitor discharging (water potential rising to zero) through a resistor and using Equation (2) by fitting an exponential curve through the first 20 min (15 to 20 data points) of Ψ_stem relaxation data following rewatering as follows (Brodribb and Holbrook 2004; Bourbia et al. 2021):

$$K_{\mathrm{r}}=C_{\mathrm{plant}}*\mathrm{l}{\mathrm{n}}^{({\mathrm{\Psi }}_0/{\mathrm{\Psi }}_{\mathrm{final}})}/t$$ (2)

where Ψ₀ was the initial water potential before rehydration, Ψ_final was the water potential after rehydration for t = 20 min, and C_plant is whole plant capacitance (mmol m⁻² MPa⁻¹; mmol g⁻¹ MPa⁻¹).

Calculating whole-plant capacitance (C_plant)

The calculation of K_r required the knowledge of whole-plant capacitance (C_plant). C_plant was obtained for 8 plants for each species (4 +AM plants, 4 −AM plants) used for K_r measurements by simultaneously measuring Ψ_stem and whole-plant mass during a dry-down from 0 to −2 MPa. At the end of the rehydration experiment, plants were removed from the pots and their roots were gently rinsed to remove all soil. An optical dendrometer was attached to the tomato petiole and pea tendril to monitor Ψ_stem during drying. The roots were carefully dried with absorbent paper, then the plants were placed onto a balance (model: VIBRA ALE-6202, 6,200 G × 0.01 G) and their mass (g) and Ψ_stem were recorded simultaneously at 10-min intervals during drying. Plants were allowed to dry slowly in the dark until their Ψ_stem reached −2 MPa (∼1 d), which corresponded to the lowest water potential the plants experienced during the drying experiment for Kᵣ measurements. C_plant was determined from the slope of the linear relationship between Ψ_stem and mass measured throughout drying, and it was normalized by total plant projected leaf area (m²) as well as root fresh weight (g). Here, the root fresh weight of the capacitance samples was determined through a regression analysis between leaf area and root fresh weight from the rehydration samples (Supplementary Figure S8).

AM colonization assessment at final harvest

After harvesting the plants from the experiment for K_r measurement, the roots of the experimental samples were stained with the ink (LAMY, Australia) and vinegar (v:v = 1:19), as detailed in Vierheilig et al. (1998). The root mycorrhizal colonization was assessed based on the methodology of McGonigle et al. (1990). Specifically, for each plant, 150 intersects were examined from 25 root segments selected randomly. The arbuscules, vesicles, and intraradical hyphae were assessed individually. The total mycorrhizal colonization was calculated as the percentage of intersects containing any of these fungal structures. The frequency of arbuscules was assessed based on the percentage of intersects containing arbuscules.

Statistical analysis

Linear regressions were used to quantify the correlation between the width variation of tomato petiole or pea tendril and Ψ_stem for each species used for K_r measurements. To assess the effects of arbuscular mycorrhizal symbiosis and Ψ_stem on K_r under drought conditions, a model selection was applied using the MuMIn package in R software (version 3.5.3; R Core Team 2024). The dredge function was used to generate a set of models, ranking them based on delta Akaike's Information Criterion corrected (ΔAICc) for small sample sizes with the formula as follows:

$$K_{\mathrm{r}}\sim (-{\mathrm{\Psi }}_{\mathrm{stem}})+\mathrm{group}+(-{\mathrm{\Psi }}_{\mathrm{stem}})*\mathrm{group}$$ (3)

where group represents different treatments in each separate experiment (wild-type tomato cultivar M82 +/−AM; wild-type tomato cultivar 76R/nonmycorrhizal tomato mutant rmc; wild-type pea cultivar Frisson/nonmycorrhizal pea mutant sym19). The best models selected for each normalization method in tomato and pea were identified based on the lowest AICc values, with models having ΔAICc ≤ 2 considered to have substantial support (Supplementary Table S1) (Symonds and Moussalli 2011). This approach allowed for a comparison of multiple models that included different combinations of −Ψ_stem, group (mycorrhizal status), and their interaction, ensuring the best model is selected based on their relative explanatory power.

Under hydrated conditions, the normality of K_r dataset was assessed for each genotype, both with and without mycorrhizal colonization. If both groups followed a normal distribution, independent t-tests were used to compare K_r between mycorrhizal and nonmycorrhizal plants. For any group that deviated from normality, Mann–Whitney U tests were applied instead. Similarly, C_plant and physiological parameters, including projected leaf area, shoot fresh weight, root fresh weight, and ratio of shoot-to-root, were compared using either t-tests or Mann–Whitney U tests, depending on the outcome of normality assessments. All statistical analyzes were performed using R software (version 3.5.3; R Core Team 2024).

Author contributions

E.F., T.J.B., I.B., and J.S. conceived and designed this study. E.F. and T.J.B. managed this project. J.S. performed the experiments and data analysis with the assistance from I.B. J.S. wrote the manuscript with the contribution from E.F., T.J.B., and I.B.

Supplementary material

The following materials are available in the online version of this article.

Supplementary Figure S1. Box plots of physiological parameters of Solanum lycopersicum L. and Pisum sativum L. plants colonized by arbuscular mycorrhiza (+AM) or with no colonization with AM (-AM).

Supplementary Figure S2. Relationship between the percentage of total root colonized by arbuscular mycorrhiza (AM) and root hydraulic conductance (K_r) normalized by projected leaf area across three genotypes in Solanum lycopersicum L. and Pisum sativum L.

Supplementary Figure S3. Relationship between the percentage of the root colonized by arbuscules and root hydraulic conductance (K_r) normalized by root fresh weight across three genotypes.

Supplementary Figure S4. Relationship between the percentage of total root colonized by arbuscular mycorrhiza (AM) and root hydraulic conductance (K_r) normalized by root fresh weight across three genotypes in Solanum lycopersicum L. and Pisum sativum L.

Supplementary Figure S5. Box plots of capacitance normalized by projected leaf area and root fresh weight across genotypes in Solanum lycopersicum L. and Pisum sativum L.

Supplementary Figure S6. Changes in petiole width and stem water potential (Ψ_stem) dynamics using optical dendrometers in Solanum lycopersicum L. during soil drying and rehydration.

Supplementary Figure S7. Calibration linear regression between stem water potential (-Ψ_stem) and petiole width (in pixels) from optical dendrometer measurements in Solanum lycopersicum L. and Pisum sativum L.

Supplementary Figure S8. Relationship between projected leaf area and root fresh weight in Solanum lycopersicum L. and Pisum sativum L.

Table S1. Model selection results for analyzing root hydraulic conductance (K_r) normalized by leaf area and root fresh weight in response to stem water potential (-Ψ_stem) in different genotypes of Solanum lycopersicum L. and Pisum sativum L.

Funding

E.F., T.J.B., I.B., and J.S. acknowledge support from the Australian Research Council including the Centre of Excellence for Plant Success (grant number: CE200100015). J.S. acknowledge support from the University of Tasmania through the Tasmanian Graduate Research Scholarship.

Data availability

The data support the findings of this study are available from the corresponding author upon reasonable request.