Arbuscular mycorrhizal fungi (AMF) are obligate biotrophic fungi that establish symbiotic relationships with plant roots (Smith and Read, 2008). The association provides benefits for plant growth and enables the AMF to complete their life cycle (Smith and Read, 2008). AMF absorb water and nutrients from the soil and transfer them to plants, while plants provide photoassimilates to fungi (Bi et al., 2005). The mycelium formed by the AMF outside the roots allows the plant to explore a larger volume   of  soil,  and this  extension  of  the  root  system increases the intake of water and nutrients (Smith and Read, 2008).

The symbiotic relationship between fungi and plants provides a number of benefits to surrounding plants and the environment (Jeffries et al., 2003; Berruti et al., 2016). Phosphorus (P) supply is the most important feature mediated by AMF (Smith and Smith, 2011). This benefit is ultimately observed primarily due to the low mobility of P in the soil (Smith and Read, 2008). AMF can contribute   to   plant   growth  through  increased drought tolerance (Garg and Chandel, 2010), increased supply of other nutrients such as nitrogen, potassium and calcium (Sharif et al., 2011; Dania et al., 2013; Crespo, 2015). They also contribute to the formation of soil aggregates by the production and release of glycoproteins such as glomalin (Rillig, 2004). Arbuscular mycorrhizal fungi interfere with plant growth and soil biology the same way chemistry and physics influence the establishment of symbiosis between plants and AMF. The availability of P in soil is linked to the production of signals and responses between plants and AMF species, directly related to soil characteristics (Smith et al., 2011; Cely et al., 2016).

Around 300 species of AMF have been formally described (Glomeromycota Species List, 2019). Among those, Rhizophagus intraradices (N.C. Schenck & G.S. Sm.) C. Walker & A. Schüßler (until recently classified as Glomus intraradices) is a generalist species found distributed throughout different environments. The list of plant species benefited by AMF on growth and productivity include crops of agricultural importance such as soybeans (Spagnoletti and Lavado, 2015), corn (Guo et al., 2014), beans (Tajini et al., 2012), wheat (Ardakani et al., 2011) and cotton (Orak and Demir, 2011). Because of these characteristics, the use of that species for the formulation of commercial inoculants is recommended.

The production of AMF inoculants bumps into the obligatory symbiotic character of these fungi. This makes it difficult to develop an efficient inoculant production system that delivers large volumes of propagules in a short period. Traditional models of multiplication of these fungi use trap plants or root tissue culture (Vosátka et al., 2012; Berruti et al., 2015). Records of an efficient production process capable of supplying the Brazilian agricultural market with arbuscular mycorrhizal inoculants have not been found in the literature.

Farmers have increased their interest in sustainable practices, and this includes introducing, maintaining, and increasing AMF populations in production systems. The application of AMF in agriculture in annual crops, which represent large areas of agricultural production in Brazil (IBGE, 2019) should be studied. Thus, considering that the arbuscular mycorrhizal species R. intraradices has the potential to assure high yields in agricultural production systems with more sustainable management, studies are needed to prove these effects.

Thus, the objective of this work was to evaluate the agronomic efficiency of a commercial inoculant based on the AMF R. intraradices, in combination with different levels of phosphate fertilization, on soybean (Glycine max L.) growth and yield in five locations with different edaphoclimatic conditions in Brazil.

Experiment locations

The experiments were conducted in the 2016/2017 crop season (September 2016 to March 2017), considering the soybean planting window, cultivars and driving practices of each tested site. For this, five (5) representative sites of soybean cultivated areas in the country (Figure 1) and with distinct edaphoclimatic characteristics were selected: Padre Bernardo in Goiás (GO), Ritápolis in Minas Gerais (MG), Tangará da Serra in Mato Grosso (MT), Pitangueiras in Paraná (PR), and Cachoeira do Sul in Rio Grande do Sul (RS) (Table 1).

Experimental specifications

Aiming to validate the efficiency of the AMF-based inoculant, the experiments were conducted in the field, following the protocols defined by MAPA (Brazilian Ministry of Agriculture, Livestock and Supply) for product registration based on the plant growth-promoting microorganism specifications, specifically the regulations contained in the IN SDA 13, from 03/25/2011 (Brasil, 2011) and the IN SDA 53, from 10/24/2013 (Brasil, 2013). Among the main requirements for the registration of microbial inoculants, we highlight the need to prove agronomic efficiency in at least four (4) regions with distinct edaphoclimatic characteristics, and that the role of the inoculant in reducing fertilization for the crop is proven.

All fertilization procedures, based on soil analysis and expected yield (SBCS, 2004), and crop treatments (Table 2) followed the recommendations for each site. The experiments were conducted until the complete pre-maturation cycle of the grains was completed, that is, in the R7 stage (Oliveira et al., 2016) (Table 1).

Crop treatments performed in each area and region are described in Table 2. For one of the evaluated states (RS), seeds were treated with specific fungicides before planting. In addition to seed treatment, fungicides were also used throughout crop development in the state of RS, as specified in Table 2. The choice of products was based on the history of fungal diseases in each evaluated site. Weed and insect control was carried out using specific herbicides and insecticides for each situation.

Evaluations

Shoot dry biomass yield, P biomass content and accumulation, and grain yield was evaluated as required by MAPA IN SDA 13 and IN SDA 53. At the phenological stage R7 (Oliveira et al., 2016), 50 plants were collected from the four central lines of each plot (working area of 10 m²). Drying of the biomass was performed in a forced air circulation oven at 60 °C to constant weight. Grain yield was determined from the yield obtained for 30 plants evaluated per plot at harvest time (considering the area mean stand), which were oven-dried following the same procedure described above.

The contents and accumulations of P in the tissues were determined from a sample composed of 20 newly mature leaves (without the petioles) harvested from the working area of each plot at the end of flowering (phenological stage R2) (Oliveira et al., 2016). Determinations followed Tedesco et al. (1995). Data was tested for homoscedasticity (Bartlet) (p≤0.05). Two-way analysis of variance was performed and Tukey’s test was used for mean separation at 5 and 10% probabilities, as established by the legislation for registration of inoculants (Brasil, 2011, 2013). The analysis was performed using the software SISVAR v.5.3.

Table 3 shows the effects of mycorrhizal inoculation, phosphate fertilization and the interaction of these factors on biomass and grain yield, along with the levels and foliar accumulation of P. In general, the mycorrhizal inoculation and phosphate fertilization significantly influenced the afore-mentioned variables, from which the main effects are presented and discussed below.

There was an increase in biomass yield linked to the increase in phosphorus supplied  by  fertilization  in  three (3) of the five (5) sites tested. In Ritápolis (MG), a site with low soil available P, the response to soybean phosphate fertilization was clearer (Figure 2e).

Inoculation provided a 29% average increase in soybean aerial biomass yield, with a range of 17 to 41% increase. At all phosphate fertilization levels, the biomass of the inoculated plants was equal to or higher than that of the non-inoculated plants. Inoculation increased biomass yield in three (3) of the five (3) areas with 0% P (average 33% increase), in all areas in the treatment with 50% of the recommended P (average 32%), and in four (4) of the five (5) areas with 100% of the recommended phosphate fertilization (average 25%) (Figure 2).

The average P contents (Figure 3) in plant tissues at the five (5) sites were 1.13 and 1.70 g.kg^-1 in the non-inoculated and inoculated plants, respectively. The highest mean content was observed in MT (2.33 g.kg^-1) (Figure 3a), and the lowest value in GO (0.79 g.kg^-1) (Figure 3d), two sites with medium P content available in the soil. In four (4) of the five (5) areas there were positive responses of P content to inoculation, except in Pitangueiras, in the State of PR (Figure 3b). The average increase coming from the inoculation on the biomass P content compared to the non-inoculated plants was 72%. In MG and RS, sites with very low and very high levels of available soil P, respectively, positive responses to inoculation were seen, regardless of the applied P dose (Figure 3e and 3c), with average increases of 84% for MG and 190% for RS.

P accumulation (Figure 4) was higher in response to inoculation in all sites with no phosphate fertilization (average increase of 155%), in four (4) of the five (5) sites with 50% of the recommended dose (average increase of 297%), and in four (4) of the five (5) sites with 100% of the recommended dose of P (average increase of 92%).

Grain yield varied in response to AMF inoculation. Grain yield was on average 29% higher in the inoculated versus the non-inoculated plants, reaching 65% increases in some locations (Figure 5). In MT, the average yield increase of the inoculated plants was 28% (Figure 5a). It was 46% for PR (Figure 5b), 16% for RS (Figure 5c), 15%  for  GO  (Figure 5d),  and  40%  for  the state of MG (Figure 5e). Inoculation provided grain yield increases at two (2) of the five (5) sites where no phosphate fertilization was applied (average increase of 34%), in three (3) of the five (5) areas with 50% of the recommended dose of P (average increase of 35%), and in four (4) of the five (5) areas with 100% of the recommended P (average increase of 22%) (Figure 5). Grain yield responded to different rates of P in three of the tested areas, with higher intensity in Ritápolis (MG) (Figure 5e), the site with the lowest initial P content available in the soil.

Comparing the average yields of the three P levels with the non-inoculated treatment at 100% of the recommended P dose, inoculation provided grain yield increases of up to 29% (Figure 5). Grain yield of the inoculated plants was equal to or greater than the non-inoculated plants that received 100% of the P dose recommended. This was true to almost all sites and P dose treatments, the only exception being the 0% P inoculated treatment in MG, a site originally with low P available in the soil.

In most situations, there was a positive response to the application of  the  tested  AMF  inoculant.  The  literature provides substantial data on the beneficial effects of R. intraradices. This species is considered generalist, not only being associated with soybeans, but also with many other plant species of agronomic interest. Among the species that are known to benefit from interaction with R. intraradices are maize, beans, wheat, rice, oats and barley (Guo et al., 2014; Spagnoletti and Lavado, 2015; Tajini et al., 2012; Ardakani et al., 2011). Species of the genus Rhizophagus (formerly classified as Glomus) are cosmopolitan, and are present practically worldwide (Davison et al., 2015), and like other microorganisms such as Rhizobium and Bradyrhizobium, generally bring remarkable benefits when associated with plant species.

The effects of AMF on plant association may vary with the environmental conditions to which plants are exposed, such as higher or lower P availability in the soil. As observed in the present study (Figure 2), Cely et al. (2016), in a study conducted in the State of Paraná, showed that plants fertilized and inoculated with Rhizophagus clarus, a physiological species very close to R. intraradices, showed biomass increases of 24% when compared to fertilized and non-inoculated plants. This increase is very similar to that observed in the present work, with an average of 25% in fertilized (100% P) and inoculated plants. Thus, it can be stated that the inoculation promoted higher biomass yield, due to the greater efficiency of the inoculant, approaching the  yields observed for grain.

For normal plant growth during the growing season, Marschner (1995) shows that the P content should be between 3.0 and 5.0 g P kg^-1 in the tissues, being toxic when exceeding 10 g P kg^-1. Some studies with soybean varieties have observed average levels that fall within this range (Corrêa et al., 2004; Broch and Ranno, 2012). However, in the present work, the average levels did not exceed 2.33 g of P kg^-1 in the dry biomass. Rezende et al. (2009) also observed values lower than those described by Marschner (1995), with minimum and maximum contents of 0.97 to 1.23 g kg^-1 of P. Thus, it is believed that although there are tables in the literature describing ranges for P values for soybean, these values may vary depending on the cultivation conditions and soybean cultivars evaluated.

In MG, a site with low available P, the effect of the inoculation on P content was clear (Figure 3e), supplying the P demand under low availability conditions. However, in places with higher P availability, such as in the State of RS, there seems to be a stimulus for extra P consumption. Even in the plots that did not receive phosphate fertilization, the P content is higher than in the control treatment (non-inoculated with 100% of P). The scientific literature highlights the potential of plant association with AMF in soils with different P concentrations available on plant nutrition and development (Harley, 1989; Bolan, 1991; George et al., 1995; Harrison, 1995; Bago et al., 2002; Ohtomo and Saito, 2005; Hodge et al., 2010; Smith et al., 2011; Cely et al., 2016). Thus, it can be seen that plants established in soils with high P content may respond differently to AMF inoculation compared to plants from low soil P available.

The largest increases in P accumulation in biomass were in RS (Figure 4c). This may have been caused by concentration as a function of higher P contents, since biomass yield did not follow these same trends. In addition, there appears to be a close relationship between AMF inoculation, biomass yield and P accumulation, especially in locations with less P availability in the soil, as there was a higher P accumulation in the five areas tested when little P was available in the soil (Figure 4).

Grain yield is the most important variable for farmers, and yield increases may be a result of different benefits provided by the association with AMF (Rilling et al., 2002; Ryan and Graham, 2002; Ruiz-Lozano, 2003; Hart and Forsythe, 2012).

Higher biomass yield, stimulated or not by inoculant application, seems to be related to higher grain yield. In the present work, grain yield of soybean plants responded to phosphate fertilization (Figure 5). The highest grain yields were observed in the places with high natural P availability in the soil: 3.8 and 3.74 Mg ha^-1 in RS and PR (Figure 5b and c), respectively, followed by MT and GO with 2.88 and 2.90 Mg ha^-1 (Figure 5a and d), respectively, and lastly MG with 2.2 Mg ha^-1 (Figure 5e), a site with very low natural P availability. However, in addition to responding to P doses, grain yield was stimulated by inoculation, and the average increase of inoculated plants over non-inoculated plants, in the five sites tested, was 29% (ranging from 15 to 46%) (Figure 5). This result proves the efficiency of the R. intraradices mycorrhizal inoculant for increasing grain yield for soybean.

These results corroborate other studies in which two (Meghvansi and Mahna, 2009) and four (Meghvansi et al., 2008) soybean cultivars were employed. The authors verified increases in seed weight when plants were inoculated with R. intraradices, proving once again the economic potential of using AMF-based inoculants for this crop. Considerations on other benefits provided by AMF in soybean were made by Porcel et al. (2003), who claim that inoculation with AMF mitigates, for example, the effect of water deficiencies on this crop.

Responses of AMF inoculation are dependent on crop type, area history and interrelationships between living communities and environmental factors. The results obtained here show that P-based controlled fertilization and the addition of R. intraradices-based inoculant resulted in a significant increase in grain yield for soybean. In addition to the direct effect on growth and yield, AMF may allow reductions in P rates applied to the soil, since it was observed that with 50% of the P dose, inoculated plants had the same grain yield as non-inoculated plants with 100% of the recommended P. Thus, this reduction has a direct effect on soil quality, reducing the dependence on phosphate fertilizers, as well as reducing the expenses related to the purchase of fertilizers by the farmer.

The R. intraradices mycorrhizal inoculant increased the biomass yield, P uptake and soybean yield under different edaphoclimatic conditions, with more marked effects on soils that originally had low or medium levels of available soil P. On average, inoculation provided an increase of 29% in biomass (regardless of the applied P dose) and grain yield, and higher P uptake. Grain yield of the inoculated plants was equal to or even greater than the non-inoculated plants that received 100% of the recommended P. This result suggests a better use of farmer’s monetary resources, as well as indicates that the use of environmentally sound approaches, such as biological inoculation of plants, cannot only reduce farms inputs of fertilizers, but also contributes towards more sustainable farming.

The authors have not declared any conflict of interests.

The authors are grateful to Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) for the scholarship to SCGS and Novatero for financial support during field trials.