Experiment conditions

The experiment was carried out in a greenhouse at the Faculty of Agronomy and Veterinary of the UASLP. As a substrate for plant growth, a mixture of soil and river sand was prepared in a 1:1 ratio. The soil used comes from an experimental area of the faculty; the sand was extracted from a riverbed. One kilogram of substrate was added to each nursery bag whose dimensions were 20 x 12.5 cm in height and width respectively. The prepared substrate presented 35, 66.3, 3 300 and 540 mg kg^-1 of P, K, Ca and Mg respectively and 34.5 mg kg^-1 (0.7% of the bases) of Na, 1.92% of organic matter (OM) as well as a pH= 6.9.

The methods used to determine these values were as follows: pH (H₂O) by potentiometer (1:2.5, soil:water ratio) (^{NC ISO 10390, 1999}), organic matter by the Walkley and Black method (^{NC 51, 1999}), assimilable phosphorus (P) by the extraction method with H₂SO₄ 0.05 mol L^-1, Ca and Mg by complexometry and Na and K by flame photometry (^{NC 52, 1999}).

According to the values of the chemical analysis, the substrate had low concentrations of K and OM, high values of assimilable P, Ca and Mg, an acceptable percentage of Na and pH close to neutrality (^{Paneque and Calaña, 2011}). The substrate had an average HMA spore number of 140 spores 50 g^-1.

Experimental design

The experiment was designed with 13 treatments, which consisted of the inoculation of two strains of HMA (HMA1 and HMA2) and three isolates of rhizobia (R1, R2 and R3) inoculated individually and in combinations, plus a non-inoculated control and a fertilization treatment as a reference. The treatments were made up as follows: non-inoculated control; single inoculation treatments (HMA1, HMA2, R1, R2, R3); combined inoculation or co-inoculation treatments (HMA1+R1, HMA1+R2, HMA1+R3, HMA2+R1, HMA2+R2, HMA2+R3), control without inoculation with fertilization (Fert). The treatments were distributed in a completely randomized design with five repetitions. The experimental unit was made up of each bag with a plant of P. laevigata.

Inoculants used

The inoculum HMA1 [IMCAM 4 Glomus cubense (Rodríguez & Dalpé)] is from the collection of arbuscular mycorrhizal fungi of the National Institute of Agricultural Sciences (INCA, for its acronym in Spanish) in San José de las Lajas, Mayabeque, Cuba. The inoculum HMA2 [Claroideoglomus claroideum (Schenck & Sm.) Walker & Schüßler] was isolated from a plot in the peanut-producing area of Ciudad Fernández, SLP, Mexico (22° 00’ 47.36” north latitude 100° 21’ 14.66” west longitude), from a soil whose textural characteristic are sandy and low clay content. Both were prepared from spores extracted from inoculums previously multiplied in a clay substrate and autoclaved at 120 °C for one hour for three consecutive days; Sorghum vulgare L. was used as a trap culture.

The culture of rhizobia cells was prepared in liquid Yeast-Mannitol-Agar (YMA) medium at 28 °C and under agitation conditions for 24-30 h (^{Vincent, 1970}). Rhizobia were isolated from plants of Vachellia schaffneri (R1 and R2) and Leucaena leucocephala (R3). The root samples of these plants were collected in their natural habitats located in sites near the Venustiano Carranza-Pajacuaran road, Michoacan (20º 07’ 09.8” north latitude 102º 36’ 52.5” west longitude) and in the locality Tocoy in the municipality of San Antonio, San Luis Potosí (21º 38’ 19.0” west longitude 98º 52’ 15.0” west longitude), respectively.

Propagation of P. laevigata

The plants of P. laevigata (mesquite) were propagated from seed. The seeds were subjected to scarification to facilitate their germination, which consisted of immersing them in water at 80 °C for two minutes, then they were immersed in a solution of 5% sodium hypochlorite for three minutes, then they were germinated in germination trays of 200 cavities with Peat moss as substrate. When the seedlings reached between 5-6 cm in height, they were transplanted into the bags containing the soil-sand mixture and inoculated. In the fertilization treatment, starting 21 days after transplantation (dat), 0.125 g of water-soluble fertilizer 17-17-17 (vigoroso excelso) was applied to each bag every 15 days, until 210 dat. The mesquite was kept growing in the bags for 12 months, throughout the period, the application of 200 ml of water to each experimental unit on alternate days or according to the needs of the plants was maintained.

Application of inoculants

The inoculation of HMA strains (HMA1 and HMA2) was carried out by applying 0.5 ml of ringer solution (the ringer solution contained NaCl 7.5 g, KCl 0.75 g, CaCl₂ 0.1 g and NaHCO₃ 0.1 g in 1 L of H₂O) containing 60 spores in each case, directly around the roots of the seedlings. In the case of rhizobia (R1, R2 and R3), 1 ml of the inoculum was applied to the substrate in a region very close to the root crown. Each 1 ml of inoculum contained 10⁸ colony-forming units (CFUs).

Evaluation of growth variables

Plant growth was evaluated based on plant height, stem diameter and number of leaflets per plant. The plant height was measured from the substrate surface to the terminal bud (cm plant) using a graduated ruler of 40 cm in length and 1 mm in precision. The diameter of the stem (mm plant) was measured with a digital king foot at 1 cm high from the surface of the substrate. At the end of the experiment (365 dat), the root was separated from the aerial part of each plant. The samples were then placed in the oven at 65°C for approximately 72 h until constant weight values were reached. The weight of MS was recorded on a Precisa LS320M SCS balance.

Evaluation of nodulation

The number of total nodules and their effectiveness were assessed as indicated in ^{FAO (1985)}. For this, the roots were extracted from the bag separating them from the substrate and carefully washed with running water. Once the roots were cleaned, the total nodules were quantified and their effectiveness was determined by observing the internal coloration by means of a cross-section of the nodule, those that exhibited red to pink coloration were considered as effective nodules for evidencing the presence of leghemoglobin, while those of white color were considered as young nodule (non-effective nodule).

Evaluation of fungal variables of mycorrhiza

Mycorrhizal colonization was determined in roots previously washed with running water and then air-dried. For the determinations, approximately 200 mg of rootlets were weighed and dried at 70 °C and stained according to the methodology described by ^{Phillips and Hayman in 1970}. The frequency of mycorrhizal colonization, which expresses the degree of occupation of the rootlets by HMAs, was evaluated by the intercept method (^{Giovannetti and Mosse, 1980}) and the visual density or intensity of colonization according to ^{Trouvelot et al. (1986}). The quantification of the number of spores (spores 50 g) was carried out from samples of 50 g of substrate from the pots, extracting these structures by wet sieving and decanting and their observation under a microscope (^{Herrera et al., 1995}).

Statistical analysis

Data were assessed using simple classification analysis of variance. Duncan’s multiple comparison test was used in cases where there was a significant effect of the treatments, to establish the differences between the means.

To meet the assumptions of normality, the data corresponding to the variable number of nodules were transformed by the square root function and natural logarithm for the visual density. In the case of (%) of colonization and (%) of effective nodules, which did not meet the assumptions of normality even after a transformation of the data, the Kruskal-Wallis test was performed to determine the existence of differences between the means, then the Games-Howell test was performed to establish the differences between the means. The IBM SPSS Statistics statistical program for Windows version 23 (Armonk, New York, United States) was used.

Fungal and nodulation variables

The treatments studied generated significant effects (p< 0.001) on the variables that characterize the mycorrhizal structures associated with P. laevigata. Treatments inoculated with both HMA and co-inoculated had a greater number of spores than the non-inoculated control and the fertilization treatment, except when combined with the isolate R2 (Figure 1A). Mycorrhizal colonization was favored when the strains of HMA were combined with the isolates of rhizobia R1 and R3 (R1+HMA1, R3+HMA1, R1+HMA2 and R3+HMA2), standing out significantly (p< 0.001) from the rest of the treatments.

Figure 1 Effect of various inoculation and co-inoculation treatments on spore number (A), mycorrhizal colonization; and (B) and visual density (C) of P. laevigata roots. ^abcdef= values with different letters in each variable differ significantly according to Duncan’s test (p< 0.05) for the variables of spore number and (%) of visual density, and according to the Games-Howell test (p< 0.05) for the variable (%) of colonization. The bars show the standard error ± of the means. R1, R2 and R3 (Rhizobia isolates). HMA1 (Glomus cubense) and HMA2 (Claroideoglomus claroideum). Fert (0.125 g of water-soluble fertilizer 17-17-17). 

Also, both rhizobia, R1 and R3, combined with HMA2 reached higher percentages of colonization than when combined with HMA1 (Figure 1B). In the case of visual density (%), R3+HMA2 followed by R1+HMA2 were the treatments with the highest values in this variable and the other treatments inoculated with HMAs, except R2+HMA1, exceeded those that were not inoculated with HMAs (Figure 1C). The non-inoculated control and the fertilizer treatment had the lowest values of spore number, colonization and visual density (Figure 1).

The nodulation of P. laevigata was higher (p< 0.001) with the inoculation of the isolates R1 and R3 and their combinations with HMAs (R1+HMA1, R3+HMA1, R1+HMA2 and R3+HMA2), with respect to the rest of the treatments and the control. In turn, the nodulation of R3 was greater when combined with both strains of HMA (R3+HMA1 and R3+HMA2) than with its single inoculation (Figure 2A). The percentage of effective nodules behaved similarly to the number of nodules, with the difference that R1, R3 and their combinations with HMA1 and HMA2 did not have differences between them but with the rest of the treatments (Figure 2B).

Figure 2 Effect of various inoculation and co-inoculation treatments on the number of total nodules (A) and the number of effective nodules (B) of P. laevigata . ^abcde= values different letters in each variable differ significantly according to Duncan’s test (p< 0.05) for the variable of number of total nodules per plant, and according to the Games-Howell test (p< 0.05) for the variable of effective nodules per plant (%). The bars show the standard error ± of the means. R1, R2 and R3 (rhizobia isolates). HMA1 (Glomus cubense) and HMA2 (Claroideoglomus claroideum). Fert (0.125 g of water-soluble fertilizer 17-17-17). 

In this research, it was found that the inoculation of HMAs, alone or in combination with rhizobia isolates, favors the generation of mycorrhizal structures and greater colonization in P. laevigata (Figure 1). Similarly, the inoculation and co-inoculation of rhizobia, simply and combined with HMAs, produced the highest number of total and effective nodules in the species studied (Figure 2). Similar results, but only with HMA inoculation, were found by ^{Monroy-Ata et al. (2007}), when inoculating seedlings of P. laevigata with an inoculum reproduced from HMA of the rhizosphere of Bouteloa gracilis, the percentage of colonization reached values greater than 50%.

In Prosopis juliflora, results have been found that suggest that inoculation combined with isolates of rhizobia and HMA favors both the colonization and formation of mycorrhizal structures and the nodulation of rhizobia with respect to the inoculation of rhizobia individually in soils with sodium chloride (NaCl) amendments (Dixon et al., 1993). Our research is pioneering in demonstrating that the inoculation of HMA and rhizobia in P. laevigata favors the formation of mycorrhizal structures and nodulation.

The control treatment also presented spores at a similar level as other inoculated treatments (Figure 1A), which is due to the presence of HMAs resident in the substrate. Other inoculated treatments had up to twice as many spores as the control. The colonization of roots with HMAs resident in the soil is a process whose magnitude is influenced by the effectiveness of the strain according to soil and climatic conditions and the affinity between symbiont and host (^{Rivera and Fernández, 2003}; ^{João et al., 2016}).

It is possible to consider that the spores registered in the inoculated treatments corresponded in a greater proportion to the inoculum than to the residents of the soil. The above, due to the homogeneity observed in the morphological characteristics of the spores when quantifying them. However, in this study the number of spores of each type of present in the sample was not recorded. This suggests that the inoculated HMA strains were more effective in promoting the development of mycorrhizal structures than those originally present in the substrate (Figure 1). However, this assertion requires greater methodological rigor for its confirmation.

It has been shown that, by inoculating the roots of plants with effective strains, both of HMA and rhizobia, the production of fungal structures and nodulation, respectively, increases. However, there are factors that influence the effectiveness of the strains. In the case of HMAs, different species behave differently according to certain ranges of soil fertility and pH, which indicates a certain between species or strain-type of soil (^{Rivera and Fernández, 2003}; ^{João et al., 2016}).

On the other hand, although rhizobia are common inhabitants in agricultural soils, their population is usually insufficient to achieve a beneficial relationship with the legume, which makes it necessary to apply inoculants to the seed or young roots to guarantee symbiosis and with it the biological fixation of nitrogen (^{Das and Varma, 2009}).

Mesquite growth variables

Inoculation favored the growth of P. laevigata, however, the number of leaflets was higher (p< 0.001) than the control only when the isolate R3 was co-inoculated with both mycorrhizal strains (R3+HMA1 and R3+HMA2) (Figure 3A). On the other hand, both the strains of HMA and the rhizobia R1 and R3 single-inoculated, as in their combinations, produced plants with greater height, like the fertilized treatment, with significant difference over the treatments with R2 alone and co-inoculated with HMAs (R2+HMA1 and R2+HMA2), which behaved similarly to the control without inoculation (Figure 3B).

Stem diameter was also positively influenced by inoculation, with all treatments being superior to the control, except the isolate R2 inoculated individually. Treatments R1, R3, R1+HMA1, R3+HMA1, R1+HMA2 and R3+HMA2 stood out for generating plants with a larger (p< 0.001) stem diameter than the rest of the treatments (p< 0.001) (Figure 3C).

Figure 3 Effect of various inoculation and co-inoculation treatments on the number of leaflets (A) height (B); and stem diameter (C) of P. laevigata. ^abcdef= values with different letters in each variable differ significantly according to Duncan’s test (p< 0.05). The bars show the standard error ± of the means. R1, R2 and R3 (rhizobia isolates). HMA1(Glomus cubense) and HMA2 (Claroideoglomus claroideum). Fert (0.125 g of water-soluble fertilizer 17-17-17). 

Biomass production was also influenced by inoculation, but in the case of the aerial part, it was the isolate R1 and the combination R3+HMA2 those that reached higher MS values. Also, all inoculated treatments and the fertilized treatment produced more biomass than the non-inoculated control (p< 0.001), except when the isolate R2 was present, both single inoculated and combined with HMAs (Figure 4A). The strain HMA1, isolates R1 and R3 and the combinations R1+HMA1, R3+HMA1, R1+HMA2 and R3+HMA2 produced greater root biomass (p< 0.001) than the treatments with the presence of the isolate R2, and the non-inoculated control and the fertilized treatment (Figure 4B).

The increase in growth provided by inoculation that was recorded in this research partially coincides with the results found by ^{Gardezi et al. (2020}), who found that, in Xerosol and Litosol-type soils, with alkaline and neutral pH respectively, inoculation with Glomus sp., together with the application of organic matter, favored the growth of shoots and roots in seedlings of P. laevigata under greenhouse conditions. In our study, inoculated HMAs produced greater colonization and density of spores, greater height and biomass production than the control, even when inoculated individually (Figures 1, 3 and 4), which demonstrates the effectiveness of these strains in soils with characteristics similar to those of the substrate used.

Figure 4 Effect of various inoculation and co-inoculation treatments on the production of dry mass of the aerial part and root of P. laevigata at 180 dat. Values with different letters in each variable differ significantly according to Duncan’s test (p< 0.05). The bars show the standard error ± of the means. MS PA (dry mass of the aerial part). MS R (dry mass of the root). R1, R2 and R3 (rhizobia isolates). HMA1 (Glomus cubense) and HMA2 (Claroideoglomus claroideum). Fert (0.125 g of water-soluble fertilizer 17-17-17). 

In another tree legume species such as Leucaena leucocephala, ^{Quintana et al. (2014}), when using the inoculation of mycorrhizal fungi and rhizobia in an isolated and combined way, obtained higher values in the growth and biomass variables of the crop in relation to the non-inoculated control, and the combined inoculations were more effective than the single inoculations.

In this research, it can be observed that, mainly for the variables of plant height and MS, the treatments inoculated with both microorganisms in a single and combined way were superior to the control, except when it was inoculated with the isolate R2, in a single way and in its combinations with HMAs (R2+HMA1 and R2+HMA2) (Figures 3 and 4).

In this sense, it is known that most species of rhizobia experience a strain-species of legume specificity (^{Balatti, 1996}). It should be noted that the interaction between the host and symbionts inoculated with the microorganisms present in the substrate, which was not sterilized, must be taken into account. Despite this, the effect on plants is observed with inoculation and co-inoculation treatments, which shows that the strains used are effective, except R2.

The higher height and biomass production experienced by mesquite seedlings inoculated with HMA1, HMA2, R1, R3, R1+HMA1, R1+HMA2, R3+HMA1 and R3+HMA2 coincides with the colonization values, spore number and nodulation of rhizobia. In turn, the highest values in these variables were reached when both symbionts were combined (Figures 1, 2, 3 and 4). This relationship between increased fungal structures, nodulation and growth of P. laevigata plants demonstrates the benefit of tripartite symbiotic interaction (rhizobia-legume-HMA) mentioned above by several authors such as ^{Rabie et al. (2005}) and ^{Lara et al. (2019}). Tripartite symbiosis favors the development of each organism individually and together, which represents opportunities for better growth of the associated plants, and, consequently, greater productivity (^{Toro et al., 2008}).

The synergy between HMAs and rhizobia in favor of plant productivity has also been corroborated by authors such as González et al. (2012), who, when evaluating the effect of co-osculation of rhizobia isolates and a strain of HMA (G. cubense), in Kudzu plants (Pueraria phaseoloides), obtained increases in yields and nutritional content of grass in the treatments where both microorganisms were combined.

The control treatment established in this investigation resulted in the presence of rhizobia nodules in their roots. Evidently, such nodules were produced by the native rhizobia or residents of the soil present in the substrate. Like any legume, nodulation of P. laevigata occurs naturally in several types of soils.

Although there is no information related to nodulation by native rhizobia in P. laevigata, the isolation of these microorganisms in other species of the genus such as Prosopis juliflora and Prosopis alba has been reported by ^{Kulkarni and Nautiyal (1999}) and ^{Velázquez et al. (2001}) respectively.

However, this research is the first to document the beneficial effect of dual symbiotic association on P. laevigata in pH-neutral soils. This evidences the effectiveness of the isolates and strains tested in this study.

It is interesting that these isolates presented compatibility with P. laevigata, even in the case of R1 and R3, they proved to be almost as efficient in some of the variables measured as their combinations with HMAs (R1+HMA1, R1+HMA2, R3+HMA1 and R3+HMA2). Our results provide new rhizobia isolates compatible with P. laevigata and that are good growth promoters, which increases the possibility of generating inoculants that can promote good results in other species, even in crops.