Dependency and Response of Apuleia leiocarpa to Inoculation with Different Species of Arbuscular Mycorrhizal Fungi

Inoculation with arbuscular mycorrhizal fungi (AMF) is a strategy to improve the efficiency of forest plantations, reducing costs and increasing the survival of plant species. The objective of this study was to assess the response and mycorrhizal dependency of seedlings of the forest species Apuleia leiocarpa (Vogel) J.F. Macbr to inoculation with AMF. The experiment was conducted in a completely randomized design using a 5 × 5 factorial arrangement with six replications. The treatments consisted of combinations of five P rates (0, 24, 71, 213, and 650 mg kg) with five types of inoculations with AMF (inoculation with the fungi Rhizophagus clarus, Gigaspora margarita, Dentiscutata heterogama, inoculation with an AMF mix of these three species, and a treatment without inoculation). The A. leiocarpa showed the highest biomass accumulations in inoculation with D. hetorogama combined with the P rates of 213 and 650 mg kg, and in the AMF mix combined with the P rates of 71, 213, and 650 kg. Biomass accumulation showed a linear, positive response to inoculation with D. heterogama combined with the different P rates, and a positive square root fit to inoculation with the AMF mix. The plants inoculated with G. margarita had no significant biomass accumulation. The plant species had a positive response to inoculation with R. clarus combined with the lowest P rates; however, it had a negative response to combination with the highest P rate (650 mg kg). The relative benefit of inoculation with these fungi was more than 100 % in most treatments, showing the high mycorrhizal dependency of A. leiocarpa and the nutritional benefit of AMF inoculation for this species. However, this response is dependent on the fungus species that colonize the plant roots. The best combination between fungus and P rate was inoculation with the AMF mix combined with the P rate of 71 mg kg.


INTRODUCTION
The inoculation of seedlings with arbuscular mycorrhizal fungi (AMF), which are the symbionts most commonly associated with higher plants (Smith and Read, 1997), contributes to successful revegetation of degraded areas and reduces costs by accelerating plant development (Franco and Faria, 1997;Chaer et al., 2011).These fungi occur in most terrestrial ecosystems and contribute to the productivity and competitiveness of their host plants (Sieverding et al., 1991), thus favoring plant diversification (van der Heijden et al., 1998).Arbuscular mycorrhizal fungi can exploit a more extensive volume of soil than root systems due to the length of their hyphae, contributing to increased nutrient and water uptake by plants.Therefore, mycorrhizal growth on plants is considered one of the most important associations between plants and microorganisms (Wang and Qiu, 2006).
The nutritional benefits and the response of plants to AMF colonization, as shown by their increased growth and productivity (Rocha et al., 2006), may vary with the host plant, the AMF species, and the environmental conditions (Rogers et. al., 1994;Saggin-Júnior et al., 1994;Sanders et al., 1996).There is evidence for functional specificity between these symbionts (Bever, 2002), which may depend on the balance between the benefits and costs of the symbiosis to the host plant (Koide, 1991) and vary due to differences in colonization rates or in nutrient transport efficiency between the myco-symbiont and the phyto-symbiont (Abbott and Robson, 1982).
Testing the inoculation of different AMF in the same plant species is needed in order to select the most efficient strains and to maximize the benefits of the symbiosis to the host plant (Saggin Júnior and Siqueira, 1995).Moreover, it is important to evaluate the plant's mycotrophic potential, which is defined as the plant's degree of mycorrhizal dependency.The degree of mycorrhizal dependency is related to the concentration of phosphorus in the soil solution.The main contribution of AMF to plants is through P uptake due to the low mobility of this nutrient in the soil and due to the high amounts needed for plant growth and metabolism (Siqueira and Saggin-Júnior, 2001).The higher the P rate at which AMF inoculation benefits the plant, the higher the mycorrhizal dependency of the plant (Janos, 1988).
Despite the findings reported in the literature, the benefits of mycorrhizal symbiosis and the factors responsible for differential mycorrhizal dependency are not well understood.This is noticeable especially in tree species of the Brazilian Atlantic Forest, for which there is only incipient information regarding their association with mycorrhizal fungi.One of these species is Apuleia leiocarpa (Vogel) J.F. Macbr (Fabaceae), a pioneer species popularly known as "garapá" or "grápia".A. leiocarpa is ecologically important due to its usefulness in recovery of degraded soils and economic contribution from the value of its timber (Carvalho, 2003).Despite being a legume, A. leiocarpa does not fix N, but it is able to associate with mycorrhizal fungi.
The hypothesis was that Apuleia leiocarpa has distinct responses to inoculation with different species of AMF.The objective of this study was to assess mycorrhizal dependency and the response of Apuleia leiocarpa (Vogel) J.F. Macbr to inoculation with arbuscular mycorrhizal fungi (AMF).

MATERIALS AND METHODS
The experiment was conducted in a greenhouse using a completely randomized design in a 5 × 5 factorial arrangement with six replications.The treatments consisted of combinations of five P rates (0, 24, 71, 213, and 650 mg kg -1 ) applied to the soil using potassium monophosphate (KH 2 PO 4 ), with five types of inoculations with arbuscular mycorrhizal fungi (AMF), namely, inoculations with the fungi Gigaspora margarita Becker and Hall (GM), Rhizophagus clarus Becker and Gerdemann (RC), and Dentiscutata Rev Bras Cienc Solo 2017;41:e0160174 heterogama (Nicol and Gerd) Walker and Sanders (DH), inoculation with an AMF mix of these three species (GRD), and a treatment without inoculation.An absolute control was used, which was formed from combination of the treatment without inoculation with AMF and a 0 rate of P.
The P rates (0, 24, 71, 213, and 650 mg kg -1 ) applied to the soil were established based on remaining P (Alvarez V et al., 2000).The K concentration of the different treatments was balanced with application of KCl.The available P in the soil was determined by the Mehlich-1 extractant, which showed that the applied P was highly correlated with the available P (Figure 1).
The substrate of each container was inoculated with 1 g of soil and roots colonized with propagules of AMF according to each treatment.The fungi species were inoculated with different amounts: Dentiscutata heterogama (DHET A2 -CNPAB002) with 17 spores g -1 , Gigaspora margarita (GMAR A1 -CNPAB001) with 35 spores g -1 , and Rhizophagus clarus (RCLA A5 -CNPAB005) with 26 spores g -1 .Ten ml of a filtrate solution from the three AMF inocula, free of propagules, was added to all containers to standardize the treatments regarding the other components of the inoculum microbiota.
The A. leiocarpa seeds were sanitized with an H 2 O 2 (30 %) solution for 2 min and immersed in an H 2 SO 4 (98 %) solution for 15 min to overcome dormancy.Subsequently, seeds were germinated in petri dishes with filter paper and cotton inside a germination chamber (Biological Oxygen Demand) under constant light and temperature of 28 °C for 5 days.
Three pre-germinated seeds were placed in each container and the seedlings were thinned after 15 days, leaving one seedling per container.
The seedlings were removed from the containers and separated into shoots and roots.
Colonization of the root system by the AMF was evaluated, with three replications.Fine roots (0.5 g) were taken from plants of each treatment and maintained in an ethanol solution (50 %) until the whitening and staining process (Koske and Gemma, 1989;Grace and Stribley, 1991).The percentage rate of root colonization by the AMF was evaluated through the gridline intersect method (Giovanetti and Mosse, 1980), adapted from the root length method (Newman, 1966).
The root dry weight (RDW) was also evaluated with three replications, whereas the shoot dry weight (SDW) was evaluated with six replications.Plants of each treatment were dried in a forced-air circulation oven (65 °C, 72 h) and weighed (g) on a 3-decimal-place analytical balance to evaluate the RDW and SDW.The RDW to SDW ratio (RSr) was calculated from these two variables.Both materials were ground in an electric mill and subjected to nitric-perchloric digestion (2:1) for extraction of P and K, and sulfuric acid digestion for extraction of N (Silva, 1999).The P contents were determined by colorimetry, K by flame photometry (Silva, 1999), and N by the modified Kjeldahl method.
The P benefit, mycorrhizal benefit (increase in dry weight from the fungal species), and symbiotic efficiency were assessed using response curves to the P applied, estimated by regression analysis, as described by Saggin Júnior and Siqueira (1995).
The relative increase of biomass (RI) in the forest species due to the treatments was defined as the increase in dry weight by the inoculation and fertilization compared to the treatment without inoculation with AMF and with a P rate of 0 (absolute control).This calculation was performed for all P rates.This relation was converted into percentage of relative increase of root dry weight (RI-RDW) and relative increase of shoot dry weight (RI-SDW), which was calculated by the formula RI = [(DWt -DWac)/DWac] × 100 in which DWt is the dry weight (RDW or SDW) of the treatment, and DWac is the dry weight (RDW or SDW) of the absolute control (without inoculation with AMF and with P rate of 0).
All data were transformed (Box-Cox) to achieve normalization and subjected to analysis of variance.The Scott Knott test at 5 % significance was applied to compare the averages.
Regression analysis was applied to the quantitative factors.The Sisvar 5.0 software (Ferreira, 2003) was used for statistical analyses.

Apuleia leiocarpa response to inoculation with arbuscular mycorrhizal fungi (AMF)
The species A. leiocarpa showed distinct responses to the different arbuscular mycorrhizal fungi (AMF) and P rates, as indicated by variations in plant biomass (root, shoot, and total dry weight) (Figure 2).
The total biomass of A. leiocarpa inoculated with R. clarus decreased linearly (Figure 2a).Inoculated seedlings accumulated more biomass than uninoculated seedlings with the four initial P rates (0, 24, 71, 213 mg kg -1 ); however, this was reversed at the rate of 650 mg kg -1 , in which uninoculated seedlings accumulated more biomass.Based on the Rev Bras Cienc Solo 2017;41:e0160174 estimated model, the combination of P with AMF is negative as of the P rate of 444 mg kg -1 (value T') (Figure 2a), which represents 225 mg dm -3 of available P (Figure 1).
Contrary to what was observed for R. clarus, the total biomass of A. leiocarpa inoculated with the fungi D. heterogama and G. margarita increased linearly with the P rates (Figures 2b and 2c), denoting a synergistic effect of these species with P availability.
The shoot biomass of plants inoculated with the fungus D. heterogama increased at all P rates compared to the plants without inoculation (Figure 2b).In contrast, plants inoculated with G. margarita had a weaker initial development up to the P rate of 139 mg kg -1 (98 mg dm -3 of available P), as estimated by the statistical model, with a positive response to inoculation beyond this rate (Figure 2c).
The accumulation of total biomass in response to inoculation with the AMF mix followed a square root model (Figure 2d).The dry weight of seedlings in this treatment increased from the P rate of 71 mg kg -1 , tending to stabilize as of the P rate of 213 mg kg -1 , with no significant variation up to the rate of 650 mg kg -1 .
The mycorrhizal benefit varied depending on the fungus species (Table 1).Higher mycorrhizal benefits promoted by the fungus D. heterogama and the AMF mix denote a symbiotic efficiency higher than 100 % in both treatments (Table 1).

Relative increase
Inoculation with D. heterogama combined with the highest P rate (650 mg kg -1 ) promoted the highest increase in shoot dry weight (SDW) of A. leiocarpa, resulting in an increase of approximately 550 % compared with the absolute control (without inoculation and P rate of 0).Inoculation with the AMF mix combined with the highest P rate increased shoot development by 492 % compared to the absolute control, with the highest root growth increase at the P rate of 71 mg kg -1 (Figure 3).
Plant responses to inoculation with R. clarus were different since the highest relative increase of shoot biomass was observed at the lowest P rates, decreasing with increasing P rates.A relative increase in roots was expressive only in the combination with the P rate of 0 (Figure 3).
Inoculation with the fungus G. margarita resulted in relative increases similar to those of the treatments without inoculation, at all P rates.The most expressive response was for root growth at the highest P rate, which resulted in an increase of approximately 300 % (Figure 3).
The response of A. leiocarpa to increasing P rates in treatments without inoculation was 20 % (24 mg kg -1 ), 13 % (71 mg kg -1 ), 38 % (213 mg kg -1 ), and 5 % (650 mg kg -1 ), indicating that the roots of this species may be inefficient in taking up P, due to genetic characteristics.These responses were low compared with those observed in the treatments of inoculations with D. heterogama and the AMF mix.Seedlings in the treatments without (1) GRD: AMF mix of the three arbuscular mycorrhizal fungi.
(2) Estimated by subtracting the area under the curve of the treatment without inoculation from the area under the curve of the treatments with inoculation.
(3) Estimated by the definite integral of the curve of the treatments without inoculation.
(4) Estimated by the relation between the mycorrhizal benefit and P benefit.
Rev Bras Cienc Solo 2017;41:e0160174 inoculation had the highest shoot growth at the highest P rate, reaching a growth approximately 200 % higher than that found in the absolute control (Figure 3).
The results, in general, showed a decrease in the RDW to SDW ratio (RSr) in treatments with inoculation compared with those without inoculation.The highest RSr were found in plants with the highest P rate, without AMF inoculation (2.16) and with inoculation with G. margarita (2.64), and in the treatment with a P rate of 0 and inoculated with R. clarus (3.36) (Figure 4), denoting higher root system growth.
All factors (fungus species, P rate, and their interaction) were significant at 5 % probability for the variables related to the biomass of the species evaluated (Table 2).
Inoculation with the AMF mix combined with the P rates of 71 to 650 mg kg -1 resulted in a higher increase in SDW compared with the treatments of single inoculations at these same rates.An exception was the treatment DH650, which had a similar increase in SDW (Table 2).
Apuleia leiocarpa had no response to inoculation with G. margarita, showing increases in RDW and SDW equivalent to those found in the treatments without inoculation (Table 2), except at the rate of 650 mg kg -1 , in which the plants showed a significant increase in RDW.Nevertheless, the treatment GM650 had no increase in SDW, indicating that the fungus G. margarita was inefficient in increasing SDW, showing more sensitivity to P than the other AMF.
The colonization rate of all treatments with AMF inoculation ranged from 56 % at the lowest P rate to 8 % at the highest P rate.The treatments without inoculation had no mycorrhizal colonization.
The highest increases in seedling height were observed in treatments GM650 (27.2 cm), HD71 (23.5 cm), HD213 (21.8 cm), and RC650 (21.7 cm) (Table 2).Plants inoculated with the AMF mix combined with the P rates of 0 and 24 mg kg -1 had high colonization rates; however, their height had little development.

Nutritional benefit
The plants of treatments inoculated with AMF had greater nutritional benefits compared with those of treatments without inoculation.These benefits were evident for P nutrition, with differences found only in the plant shoots (Table 3).No significant differences were observed in N and K contents in either roots or shoots.The treatments consist of letters representing the arbuscular mycorrhizal fungi species (DH: Dentiscutata heterogama, GM: Gigaspora margarita, RC: Rhizophagus clarus, and GRD: DH+GM+RC), followed by numbers representing the P rates (0, 24, 71, 213, and 650 mg kg -1 ); the numbers without letters represent the P rates with absence of fungi; the rate 0 represents the absolute control.RDW: root dry weight; SDW: shoot dry weight; TDW: total dry weight; RSr: root to shoot ratio; D: stem diameter at the soil surface.Averages followed by different lowercase letters in the columns indicate significant difference between treatments by the Scott Knott test (p=0.05).Averages followed by different uppercase letters in the columns indicate significant difference between treatments with the same P rate and different fungi species by the Scott Knott test (p=0.05).

DISCUSSION
The forest species A. leiocarpa, which is native to the Brazilian Atlantic Forest, showed greater gains in biomass when in symbiosis with arbuscular mycorrhizal fungi (AMF).The results showed that A. leiocarpa has different responses to inoculation depending on the AMF species, as also reported for other plant species (Saggin-Júnior et al., 1994;Siqueira and Saggin-Júnior, 2001;Jansa et al., 2007).The response to the increase in P rates was linear and positive for the plants inoculated with the fungi D. heterogama and G. margarita, and linear and negative for those inoculated with Rhizophagus clarus, indicating an ecological difference between these fungi species.Moreover, A. leiocarpa had a different response to inoculation with the AMF mix, whose data fitted a positive square root model with increasing P rates and showed a significantly higher biomass increase at the lowest rates, indicating synergism between the fungi.Moreover, the fungi probably colonize the roots according to their preferential P rates, thus, better supplying the plant needs under different conditions than species applied singly (Figure 2).The treatments consist of letters representing the arbuscular mycorrhizal fungi species (DH: Dentiscutata heterogama, GM: Gigaspora margarita, RC: Rhizophagus clarus, and GRD: DH+GM+RC), followed by numbers representing the P rates (0, 24, 71, 213, and 650 mg kg -1 ); the numbers without letters represent the P rates with absence of fungi; the P rate 0 represents the absolute control.Averages followed by different lowercase letters in the columns indicate significant difference between treatments by the Scott Knott test (p=0.05).
Rev Bras Cienc Solo 2017;41:e0160174 Apuleia leiocarpa can be classified as a species of high mycorrhizal dependency, considering its significant responses to inoculation with D. heterogama and the AMF mix, and the high mycorrhizal and relative benefits obtained from inoculation with these fungi (Janos, 1988;Siqueira and Saggin-Júnior, 2001).
The higher shoot growth of inoculated A. leiocarpa resulted in a lower ratio of root (RDW) to shoot dry weight (SDW) (Zangaro et al., 2005).Similar results were found in all treatments with P rates and without inoculation, and those inoculated with G. margarita, exhibiting an increase in the RDW to SDW ratio (RSr) with increasing P rates, except in the treatment inoculated with R. clarus and a P rate of 0 (CR0).This result indicates that this fungus induces the root system development of its host under low P availability.Pereira et al. (1996)  Height increases were not related to increases in plant biomass, since the plants of most of the treatments had similar heights, unlike the results reported in previous studies (Saggin -Júnior et al., 1994;Pouyú-Rojas et al., 2000).
Another positive effect of mycorrhizal inoculation was improvement in P nutrition, denoting the importance of inoculation for P uptake in A. leiocarpa.This improvement was observed in plants inoculated singly or with the mix of the three fungi species.Inoculation with AMF favored P content in the shoots, regardless of the P rates.For example, the P content of the AMF mix inoculation treatment was 6.79 times greater than the P of the plants without inoculation combined with the P rate of 0. Similar results were found by Moreira-Souza and Cardoso (2002), who evaluated the species Araucaria angustifolia inoculated with the fungus Glomus intraradices, Gigaspora rosea, and a native AMF mix, and they found increases in the amount of P taken up and accumulated with increasing P rates.According to these results, the main effect of the fungi for plant nutrition is improvement in uptake of low mobility nutrients, especially P, Cu, and Zn (Habte, 2000).
The colonization rate varied depending on the AMF species and P rates.However, the colonization of all fungi species tended to be inverse to the P rates, decreasing with increasing P rates.This result was probably because available P content is one of the main limiting factors for symbiosis (Smith and Read, 1997;Rocha et al., 2006;Parniske, 2008;Renuka et al., 2012).
The response models vary depending on the plant species since the increase in A. leiocarpa growth showed a linear fit to the fungi inoculations, whereas Saggin- Júnior et al. (1994) reported a square root fit for coffee seedlings inoculated with G. margarita and R. clarus, and negative responses from high P rates.
Functional specificity among fungi species is a possible hypothesis since inoculations with the AMF mix promote additional growth in the host plant compared with inoculations with a single fungus.Arbuscular mycorrhizal fungi show different efficiencies with the use of P, even among isolates of the same species, and different AMF species have different strategies for soil utilization and root colonization, indicating the complementarity and benefit of greater richness of species in the system (Maherali and Klironomos, 2007).This is a favorable factor for use of the AMF mix, since plants inoculated with the AMF mix were able to adapt to different environmental conditions.Studies show that greater richness and complementarity are related to greater productivities of plant communities.
These results have important implications for studies on plant-AMF interactions.For example, studies on mycorrhizal dependency must consider inoculations with different AMF species to avoid underestimating the response and dependency of the plant species Rev Bras Cienc Solo 2017;41:e0160174 under evaluation.The A. leiocarpa responses (positive or negative) to increasing P rates varied with the AMF species, indicating that the choice of the fungus species for estimating mycorrhizal dependency can lead to different conclusions.The plants that were not inoculated had very small responses to an increase in the P rates, also denoting a high dependency of A. leiocarpa on mycorrhizae for uptake of this nutrient, and implying a very high Janos' T value.Other authors have reported varied responses of this plant to AMF species (van der Heidjen et al., 1998, Siqueira andSaggin-Júnior, 2001); different responses to inoculations with the fungi R. clarus and D. heterogama were found.
Results indicate that inoculation with the AMF mix was the best treatment, to which plants showed significant responses when combined with low and intermediate P rates, and thus, it can be recommended for producing A. leiocarpa seedlings.The positive responses to lower P rates mean a decrease in fertilization costs.Other studies reported the benefits of co-inoculation of AMF, such as greater tolerance to fungicides (Schreiner and Bethlenfalvay, 1997) and functional complementarity (Jansa et al., 2007), which were not assessed in this study.
Another implication is that inoculation with the AMF mix may benefit the plant by increasing functional redundancy (Jansa et al., 2007), since different species may have different responses to environmental conditions after planting.AMF may have different soil use strategies; thus, a fungi community can benefit the plant in different conditions.The diversity in a community of mycorrhizal species with different functional responses may contribute to increase the productivity of plant communities (Maherali and Klironomos, 2007).
From the ecological point of view, these results are relevant to establishing species in a field, but raise questions about the role of soil community compositions on the plant growth of forest communities, especially in the Atlantic Forest.
All these characteristics must be considered to choose forest species and mycorrhizal fungi for recovery of degraded areas, obtaining the most appropriate and fast growing seedlings and, consequently, decreasing costs.Thus, the selection of symbionts with greater compatibility tends to facilitate the recovery of degraded areas, reduce chemical inputs, and improve inoculant production conditions by working with a smaller number of plant species and mycorrhizal fungi of greater efficiency, generating good quality seedlings.

CONCLUSIONS
The A. leiocarpa seedlings showed high mycorrhizal dependency, and seedling response to inoculation with arbuscular mycorrhizal fungi (AMF) was dependent on the fungal species colonizing their roots.Apuleia leiocarpa was highly responsive to inoculations with Dentiscutata heterogama and the AMF mix (Dentiscutata heterogama, Gigaspora margarita and Rhizophagus clarus).
Inoculation of Apuleia leiocarpa with the AMF mix under greenhouse conditions resulted in the greatest growth; thus, it is recommended for production of A. leiocarpa seedlings.The best treatment was inoculation with the AMF mix combined with the P rate of 71 mg kg -1 .

Figure 4 .
Figure 4. Comparative analysis of the ratios of RDW (root dry weight) to SDW (shoot dry weight) of Apuleia leiocarpa seedlings with the highest biomass increases due to inoculation with the AMF (arbuscular mycorrhizal fungi), Dentiscutata heterogama (DH), Gigaspora margarita (GM), Rhizophagus clarus (RC), and AMF mix of these three species (GRD) and phosphorus rates (0, 24, 71, 213, and 650 mg kg -1 ).Averages followed by different lowercase letters in the columns indicate significant difference between treatments by the Scott Knott test (p=0.05).
evaluated inoculation with the fungus Glomus etunicatum (Becker and Gerdemann) in the forest species Acacia mangium (L.), Senna macranthera (Collad.)I and B, Senna multijuga (L.C.Rich) I and B, and Anadenanthera peregrina (L.) Speg.and found different responses of these plant species, reporting increases in the RSr at the lowest P rate only for Senna macranthera (Collad.)I and B., Senna multijuga (L.C.Rich) I. and B., and Anadenanthera peregrina (L.) Speg.

Figure 1. Relationship
between applied P and available P (significant at 1%, by the F test) in the soil, evaluated through the Mehlich-1 extractant.

Table 1 .
Response of Apuleia leiocarpa to inoculation with different arbuscular mycorrhizal fungi (AMF) regarding mycorrhizal benefit, P benefit, and symbiotic efficiency, considering total plant dry weight

Table 2 .
Growth of Apuleia leiocarpa seedlings inoculated with different arbuscular mycorrhizal fungi (AMF), combined with five phosphorus rates

Table 3 .
Interaction between phosphorus rates and different arbuscular mycorrhizal fungi (AMF) on nutritional characteristics of Apuleia leiocarpa