Location and environment of the experimental site

The actual research took place (February to November 2013) at a nursery of the Experimental Station Rosario Izapa which belongs to the Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias (INIFAP), located at 18 km of the Tapachula-Cacaohatan highway in the municipality of Tuxtla Chico, Chiapas State between 14° 40’ N and 92° 10’ W at 435 masl. In this place the seeds of T. donnell-smithii seeds were collected which came from a dense crown, straight stem and without any signs of plagues or disease. The weather is humid warm in summer and with a monzoon influence. The average precipitation is 4 720 mm a year and the mean temperature, isotermal, of 25.4 °C.

Substrate physical and chemical analysis

The substrate was formed with a mixture of Molic andosol soil and washed river sand, 1: 1 (V / V), whose characteristics were sand-shallow texture(Bouyucos) (80.66 % sand, 16.36 % silt, 2.98 % clay), porosity (34 %), real density (2.09 g mL^-1); bulk density (1.38 g mL^-1); 3.1 % organic matter (Walkley-Black); 0.12 ds.m^-1, electrical conductivity; pH 5.71; 0.19 % N (Kjelddhal); 5.0 ppm P (colorimetry); 18.5 ppm K⁺⁺ (atomic spectrophotometry); 59 ppm Ca⁺⁺ (atomic spectrophotometry); 8.1 ppm Mg⁺⁺; and 5 Meq.100 g^-1 cation exchange capacity. Black plastic bags (700 gauge) were filled with the non-sterile substrate, with a capacity of 5 kg of soil per bag, and were drilled in the lower part to favor drainage.

Microoganisms and inoculation

The microorganisms that were used were Rhizophagus intraradices produced in the Rosario Izapa Experimental Station of INIFAP. The inoculum contained 40 spores per gram of sterile soil and Brachiaria decumbens L. roots with 95 % of colonization. Azospirillum brasilense was purchased at the Benemérita Universidad Autónoma de Puebla with 9 x 10⁶ bacteria g of peat^-1. Pseudomonas fluorescens Migula, was provided by Celaya Experimental Station of INIFAP, with 9 x 10⁶ bacteria g of peat^-1. Seeds were germinated in trays with the same substrate and were transplanted to pots with two real leaves. At the time of transplantation, in the sewing hole were applied 2 g of inoculum mixed according to the treatment.

The five treatments are described as follows: 1) Azospirillum brasilense + Pseudomonas fluorescens; 2) Rhizophagus intraradices + P. fluorescens; 3) R. intraradices + A. brasilense; 4) A. brasilense + R. intraradices + P. fluorescens; and 5) Control. They were used with five plants as replications in a completely at random design. The experimental unit was a pot with one plant. Watering was made with distilled water.

Variables in record

Morphological and physiological yield variables were recorded 28 days after sowing (dds); from the first sampling to 46 dds, and the following to 74, 102, 130 and 158 dds. Root colonization and N and P content in plant tissue were recorded at the end of the study.

The aerial and root dry matter of the plant was determined in a forced air oven for 72 h at 75-80 °C (Sheldon, VWR 1390FM, USA) and the performance components were weighed on a semianalitical scale (Ohaus, Adventurer Pro).

The leaf area (cm²) was obtained with a leaf area integrator (LI-COR, LI 3000A, USA).

By the staining and polishing technique of ^{Phillips and Hayman (1970)}, 100 root segments with a length of 1.5-1.6 cm were prepared from each sample and were observed under optical microscope with immersion objective (100 X). The results were expressed as percentage of mycorrhizal colonization.

Finally, the contents of N (Micro-Kjeldahl) and P were quantified in a spectrophotometer (Thermo Fisher Scientific, model 400 ¼).

Statistical analysis

The effects between treatments were estimated with an analysis of variance for the morphological, physiological variables, the nitrogen and phosphorus content with the PROC ANOVA procedure. Subsequently, a comparison of means (Tukey, P ≤ 0.05) was done with the Statistical Analysis System, version 8.1 (^{SAS, 1999-2000}).

Results and Discussion

Table 1 shows the results of the different treatments. The R. intraradices, A. brasilense and P. fluorescens combined inoculation significantly increased all the growth parameters.

Table 1 Height, foliar area and dry weight of root and stem of Tabebuia donnell-smithii Rose plants. 

dds = Days after sowing; CV = Coefficient of variation (%).

* Values with different letter within each factor and column are statistically different (p ≤ 0.05).

Plants had slow growth the first 45 dds and, consequently, reached low height. At 28 days, an increase in all treatments was observed. This dynamic was repeated in the following samplings. The significant statistical differences (P ≤ 0.05) corresponded to the 46, 102, 130 and 158 dds. Between 46 and 74 dds, treatments with the microorganisms presented higher plant height compared to the control, and the average values varied from 1.32 cm, with inoculation of the three microorganisms, to 2.42 cm in the plants treated with R. intraradices and A. brasilense.

In all cases, the next evaluation period, at 102 dds, there was an increase in plant height: 4.0 at 7.0 cm; the highest height was obtained with the inoculation of the three microorganisms and the lowest with the two bacteria A. brasilense and P. fluorescens and the control. After the third sampling and until the end of the evaluation, differences in height were greater than 2 cm in treatments with microorganisms, compared to the control.

In other perennial species such as Theobroma cacao L., ^{Aguirre-Medina et al. (2007)} indicate that at 120 dds there are average differences in plant height of 9 cm in those inoculated with R. intraradices, in relation to the control. ^{Karthikeyan et al. (2009)} cite increase in plant height by inoculating the seeds of Catharanthus roseus (L.) G. Don. with A. brasilense and P. fluorescens; however, in the present study it was the combination with lower induction of plant height.

In wheat, ^{Naiman et al. (2009)} show different effects on the use of carbon sources when Azospirillum and Pseudomonas are inoculated and indicate their effect on growth. ^{Zambrano and Díaz (2008)} cite that A. brasilense promotes greater plant height in Gmelina arborea Roxb with Glomus manihotis Howeler, Sieverding & Schenk and Paraglomus occultum (Walker) Morton & Redecker. This background suggests the contrasting functionality of co-inoculations in interaction with plants (^{Jäderlund et al., 2008}).

The leaf area evidenced growth values, statistically different (P ≤ 0.05) from the first sampling with the symbiosis of the three microorganisms. Only, at 130 dds of the A. brasilense + P. fluorescens bacteria it had a statistical difference (P ≤ 0.05). In other plants, such as cocoa and coffee with R. intraradices, there is also an increase in leaf area under nursery conditions (^{Aguirre-Medina et al., 2007}; ²⁰¹¹). In this regard, ^{Sylvia (2005)} cites that plants after mycorrhizal colonization increase their photosynthetic activity, and what is expressed in more accumulation of biomass.

Dry root and stem weight

The highest biomass accumulated in the root system and the stem during the first sampling was recorded with the treatment of the three microorganisms and in the control were also statistically different (P ≤ 0.05) to the other treatments. The slow initial growth in plants inoculated with two microorganisms together is likely to be related to the increased demand for photosyntates by microorganisms during the establishment phase, even though with the three microorganisms the demand for carbon sources is greater and, consequently, initial growth is lower. It is possible that the combination of mechanisms of action between the three microorganisms depends on the functional compatibility in the physiology and biochemistry of the interaction between the microbial components (^{Vázquez et al., 2000}).

At this stage, the availability of carbohydrates to the stem decreases, since most of the compounds are required by the root (^{Roveda and Polo, 2007}). ^{Wright et al. (2005)} documented large concentrations of soluble carbohydrates in mycorrhizal roots of maize, adding that after the period of establishment of the symbiosis, normally initiates the mechanism of transport of nutrients to the plant and the concomitant increase in biomass, as happened in the following samplings with the inoculation of fungi and bacteria.

In the samples corresponding to 102, 130 and 158 dds, the largest growth in the root system occurred when inoculating the three microorganisms together. The combination of bacteria and fungi, P. fluorescens + R. intraradices improves root growth. P. fluorescens can stimulate root proliferation (^{Sánchez et al., 2005}) through the production of indoleacetic acid (^{Patten and Glick, 2002}), which is related to the promotion of plant growth (^{Gamalero et al., 2003}). Co-inoculation with the two bacteria A. brasilense and P. fluorescens produced 30 % less radical biomass and 38 % in the control at the end of the study, compared to the treatment with the three microorganisms.

The dry weight of the stem was superior and statistically different (P ≤ 0.05) to the other treatments with the inoculation of the three microorganisms, throughout the evaluation. In two samples, at 102 and 158 dds, in the treatment with R. intraradices + P. fluorescens accumulation of biomass similar to that observed with the application of the three microorganisms was found, but without statistical difference between them (P ≤ 0.05). A similar behavior was observed during the first sampling (Table 1).

The co-inoculation with the two bacteria generated less biomass in root and stem in the samples at 74, 102 and 130 dds, compared to the treatments integrated by the bacteria in combination with R. intraradices. No colonization by the fungus (MA) at the beginning of the evaluation was obtained, but at 74, 102 and 130 dds, with an increase of around 20 %, at the end of the evaluation, the percentage was 26.

The mycorrhizal colonization might belong to the fungi in the substrate, whose presence is confirmed in addition, to the percentages of the control (without inoculation). The autoctonous fungi (MA) induced smaller vegetal development of T. donnelli-smithii and the greater vegetal response obtained from R. intraradices suggests a differential effect on the ability of fungi to stimulate plant growth, even when there is a possibility of having been colonized as well by autoctonous fungi (MA) (Figure 1).

The vertical line indicates ± the standard error of five replications.

Figure 1 Mycorrhizal colonization in the root of Tabebuia donnell-smithii Rose plants. 

The co-inoculation of R. intraradices + P. fluorescens led to a greater increase in dry biomass of root and stem, compared to the R. intraradices + A. brasilense combination. The effect was more evident from 74 and 158 dds, although mycorrhizal colonization was higher with the R. intraradices + A. brasilense combination during the whole evaluation period, when compared to the R. intraradices + P. fluorescens combination, in which there was little root colonization of the fungus (MA) at the beginning, but increased by about 60 % at the end of sampling.

The co-inoculation of A. brasilense and P. fluorescens bacteria responded differently with the same fungus (MA). This suggests some specificity between the bacteria and R. intraradices. Thus, coinoculation of the fungus with the specific bacterium may induce synergism in growth (^{Trabelsi and Mhamdi, 2013}). With A. brazilense the initial colonization was the highest and with P. fluorescens, it was at the end of the study, and the same behavior was observed with the inoculation of the two bacteria together with the fungus (MA). In wheat the root growth is increased with Azospirillum and radical colonization is not affected when inoculated with G. mosseae (Nicolson & Gerd.) Gerd. & Trappe and G. macrocarpum Tul & Tul (^{Russo et al., 2005}).

In treatments with R. intraradices, the increase in the development of the host plant may be due to the greater absorption of nutrients, through the fungus hypha. It is possible from the increase in the absorption capacity of mycorrhized plants (^{Leigh et al., 2009}), due to a stimulatory effect on mycelial development, as stated by ^{Barea et al. (1998)} with Pseudomonas sp. F113 strain and G. mosseae; in addition to nitrogen fixation by bacteria (^{Hungria et al., 2004}). Results confirm the ability of R. intraradices to colonize, and their effect on inducing higher host plant growth, in annual crops, as well as on different maize varieties in Africa (^{Wright et al., 2005}); with Creole maize, varieties and hybrids in Mexico (^{Aguirre-Medina, 2006}); and perennial crops like Coffea arabica L. (^{Aguirre-Medina et al., 2011}), Coffea canephora Pierre ex A. Froehner (^{Ibarra-Puón et al., 2014}) and Cedrela odorata L. (^{Aguirre-Medina et al., 2014}).

Mycorrhizal colonization associated with bacteria does not seem to be related to the induction of plant development of Tabebuia donnell-smithii, since it varies according to the combination of the fungus with the bacterium. However, it is expected that mycorrhizal symbiosis presents changes in colonization levels (^{Sylvia, 2005}) due to the interaction of various environmental and management factors (^{Andrade et al., 2009}).

There is evidence that mycorrhiza interacts with a wide variety of organisms in the rhizosphere, such as diazotrophs; although the demand for carbohydrates increases with the co-inoculation of more than one microorganism (^{Lugtenberg et al., 1999}), and it has been estimated that about 20 % of the total assimilated carbon by the plant is transferred to the fungus (^{Sylvia, 2005}). In the study that is documented, the increase in the accumulation of biomass in the treatment with the three microorganisms together suggests the functional compatibility of them with the host plant, and that it supplied enough carbon to the microorganisms.

N and P content

In regard to nutrients, Figure 2 shows the contents in T. donnellsmithii at 140 dds.

The vertical line indicates ± the standard error of five replicates. Different letters are statistically significant (P≤0.05). Coefficient of variation in nitrogen 6.4 % and in phosphorus 6.0.

Figure 2 Variation of nitrogen and phosphorus content in the plant tissue of Tabebuia donnell-smithii Rose by treatment at 140 days. 

The nitrogen content in the stem of the plants inoculated with A. brasilense + P. fluorescens, R. intraradices + P. fluorescens; and the three microorganisms together were higher compared to the rest of the treatments, but with no statistical difference. This result can be related to competition (^{Neuhauser and Fargione, 2004}) or incompatibility (^{Oyekanmi et al., 2007}) among microorganisms, since its functional incidence has a primary effect on plant nutrition and, consequently, on the growth of the plant (^{Cano, 2011}). With Glomus fasciculatum (Thaxt.) Gerd. And Trappe did not present increase of the nitrogen content in the plant tissue of Astronium graveolens Jacq., Tectona grandis L.f., Terminalia amazonia (Gmel.) Exell and Gmelina arborea Roxb., but it did in other nutrients (^{Hernández and Salas, 2009}). In other investigations, inoculation of R. intraradices + A. brasilense increased the N content in the plant tissue of C. arabica (^{Aguirre-Medina et al., 2011}).

The highest phosphorus content in plant tissue was recorded when R. intraradices was co-inoculated with A. brasilense and with the three microorganisms together. The concentrations were statistically different (P ≤ 0.05). Many studies have shown that mycorrhizal plants absorb P from soil more efficiently than non-colonized plants (^{Andrade et al., 2009}). In Eleusine coracana Gaertn inoculated with Glomus fasciculatum and A. brasilense, greater uptake of P is recorded compared to control plants (^{Geeta et al., 2013}). It has been reported that fungal hyphae have a higher affinity for phosphate ion when their concentration is low in the solution (^{Tajini and Drevon, 2012}), and in low input production systems, mycorrhizal activity is more effective (^{Grant et al., 2005}), with emphasis on low fertility soils (^{Mudge, 2003}).

Co-inoculation of P. fluorescens with R. intraradices did not lead to higher P content in plant tissue. It is likely that this effect is related to the functionality of the two microorganisms and their effect on the T. donnell-smithii plant.