Description of the study area

This work was developed in 2013 at the Ranch "San Lorenzo" located in Metepec, State of Metepec, at an altitude of 2606 m, located at 19° 14' 866'' north latitude and 99° 35' 240'' west longitude. The predominant climate is C (W₂) (W) (O_i) corresponding to temperate g with summer rains, average annual temperature 13 °C, with frequent frosts from October to March and an annual rainfall of 785 mm (^{Pérez et al., 2014}). The most common soil is andosol, rich in organic matter.

Genetic material

Two cultivars were used: San Pedro Tlaltizapán (C1) and Santiago Tianguistenco (C2); C1 was provided by the Research Institute and Agricultural, Aquaculture and Forestry Training of the State of Mexico (ICAMEX) and was characterized by ^{Pérez et al. (2014)}. C2 is a collection which is name by their place of origin.

Treatment structure

Vermicompost, manure and mushrooms were considered in 1, 2 and 3 t ha^-1 and the formula N60-P60-K30. 10 combinations of fertilizers and both cultivars with and without G. fasciculatum originating 40 treatments (Table 1).

Design and plot size

The 40 treatments were randomized in the field under an experimental design of randomized complete block with three replications in a split plot arrangement: both cultivars with and without mycorrhiza were assigned to big plot and the nine organic fertilizers and chemical formula formed the small plot (Table 1). The plot consisted of three rows of 9.6 m² with plants spaced at 0.40 m; the useful plot was the central row (3.2 m²).

Inoculation

For G. fasciculatum suggested 2 kg ha^-1, with 79 spores per g. 0.9 g of inoculum were weighted for each treatment, the night before seeding was applied to the seed and this were disinfected with sodium hypochlorite at 5% for three minutes and then washed with water. The "garapiñado" technique consisted of 300 g of sugar dissolved in 1 L of water; in this the seeds were immersed for 10 seconds, then drained and dusted with mycorrhizal inoculum; the samples were dried on a blanket in absence of light.

Development of experimental work

Soil preparation consisted of fallow and harrowing. Manual seeding was held on April 17 2013. Treatment 60 N-60P-30K was prepared with urea (46% N), triple calcium superphosphate (46% P) and potassium chloride (60% K). Along sowing incorporated manure, vermicompost and mushroom compost in 1, 2 and 3 t ha^-1. The chemical composition of fertilizers are shown in Table 2. On April 24 was applied an irrigation. Weeding out were held on May 13 and July 4. Weed control was manual. To prevent diseases applied 1 kg ha^-1 of Manzate (mancozeb). To reduce disease incidence applied Rogor (dimethoate) at doses of 250 g in 100 L of water, before flowering; also applied 100 L ha^-1 of Anasac (Cyhalotrin) to control aphids. Harvest was made in December 2013.

Study variables

10 plants were selected in the useful plot and recorded: days to flowering (DF), floral knots (NNF, in its central axis), plant height (AP: cm from the base of the main stem), height to the first pod (APV: cm from the base of the main stem to the base of the first pod), leaf greenness index (IV: determined with SPAD 502 plus), branches per plant (NR), pods per plant (NVP), weight pod per plant (PVP: was determined in g with a digital scale), clean seeds (NSL), stained seeds (NSM), weight of clean seeds (PSL: g), weight of stained seed (PSM, g), 100 seeds weight (PS100, g) and grain yield (RG, t ha^-1) (^{Pérez et al., 2014}).

For colonization were randomly selected three plants in each useful plot, with destructive sampling and recorded root length (LR, cm), pink nodules (NDRs, with a colony counter Q-14 Solbat), white nodules (NDBl, with colony counter Q-14 Solbat), colonization [% COL: with ^{Philips and Hayman (1970)} technique from 10 fragments stained root of 1.0 cm in length recording the presence of self-structures of the fungus MA per each optical space of microscopic observation at 400x. The number of fields in which some structure was seen was considered on the number of fields seen and multiplied by 100. Then quantified the presence of fungal structures; [Vesicles (% VES), arbuscules (% ARB) and hyphae (% HIF)]. To stain root samples the following reagents were used potassium hydroxide 10%, hydrogen peroxide 10%, hydrochloric acid IN, lacto-glycerol and blue tryptophan 0.05%.

Statistic analysis

Data were subjected to analysis of variance and mean comparison with Tukey test (α= 0.01). Arithmetic means from the 40 treatments for 21 variables were used to generate the data matrix that allowed to obtain the principal component analysis (^{Sánchez, 1995}; ^{Pérez et al., 2014}).

Variance analysis

The effects for both factors and their interaction were significant in 19 of the 21 variables (Table 3); the levels from factor A (cultivar + mycorrhiza) were significantly different (p= 0.01) in 20 variables. The ten levels from factor B (fertilizer) contributed to phenotypic differentiation (p= 0.01) of DF, NNF, AP, APV, IV, NR, NVP, PVP, NSL, PSL, P100S, RG, LRz, NDRs, NDBs,% COL, %VES, %ARB and %HIF. These results show a favorable response in Vicia faba using mycorrhizae (Gosling et al., 2006) and organic fertilizers (^{Castro et al., 2009}; ^{Álvarez et al., 2010}). A x B interaction was also significant. In another study, there was positive effect of arbuscular mycorrhizal with composted manure or crop residue (^{Saif, 1986}) by increasing the potential of mycorrhizal inoculum in the soil.

Comparison of means for factor A (cultivar + mycorrhizae) With G. fasciculatum cv. San Pedro (C1) expressed the average maximum in floral knots (15.5), plant height (111.01 cm), greenness index (49.01), pods per plant (18.45) and pod weight per plant (48.48 g). This response was similar to that reported by ^{Wu and Zou (2010)} in citrus by adding mycorrhizae. In weights of seed, 100 seeds weight and grain yield were statistically equal to control; the two cultivars had the highest seed weight per plant by adding mycorrhizae. cv. San Pedro showed significant differences (α= 0.01) in root length, pink nodules, white nodules, and colonization percentages of vesicles and arbuscules. Root length (21.50 cm) was higher than that reported by ^{Thanuja et al. (2002}; 9.95 cm), but colonization (58.58%) was lower than that reported by them (79.17%), the percentage of arbuscules was 54.05; this characteristic is important because with these values there is greater absorption of nutrients and symbiosis (Table 4).

In pepper (Capsicum annuum L.) a reduction in arbuscules (21.6%), vesicles (42.89%), and total root colonization (62.7%) was observed regarding to control by inoculating with G. fasciculatum (^{Davies et al., 2002}). ^{Dutt et al. (2013)} mentioned that it was more effective the inoculation in Prunus armeniaca L. with G. fasciculatum.^{Singh et al. (2013)} reported 71% colonization by combining G. fasciculatum and G. monteilii in Coleus forskohlii and with G. fasciculatum this was 65%. In another study the addition of G. fasciculatum resulted in an increase in N, P. K, Ca, and Mg, soluble sugars, free amino acids, proline accumulation and presence of a peroxidase and catalase in saline conditions with wheat (^{Talaat and Shawky , 2011}). Moreover sucrose increased in citrus by applying Funneliformis masseae (^{Wu et al., 2013}). Inoculation with G. fasciculatum in four forest species in field and nursery resulted in an increase of basal diameter, plant height, dry weight of foliage and root and nutrients up take (^{Hernández and Salas, 2009}).

Factor B (organic fertilizers)

Chicken manure (B3) and mushroom compost (B9) in 3 t ha^-1 showed the best phenotypic expression in floral knots (NNF), plant height (AP), greenness index (IV), pods per plant (NVP ), pod weight per plant (NVP), clean seeds (NSL), weight of clean seeds (PSL), 100 seed weight and grain yield (RG 2.42 t ha^-1) (Table 5). These results are similar to ^{Pool et al. (2000)} who obtained an increase in yield of maize (Zea mays L.) and cocoa (Theobroma cacao L.) attributable to increased phosphorus in poultry manure (^{Orozco and Thienhaus, 1997}). In potato (Solanum tuberosum L.) the addition of organic fertilizers caused increased production and quality of tuber (^{Romero et al., 2000}).

In root length (20.60) the maximum value with 1 t ha^-1 and pink nodules (20.78) with 3 t ha^-1 of chicken manure was observed. But for white nodules (10.33) the maximum value was recorded in 3 t ha^-1 mushroom compost. The highest colonization was with mushrooms in 2 t ha^-1 (40.7%). 47.87% of vesicles was associated with poultry manure at 2 t ha^-1 and 36.78% of arbuscules and 45.54% hyphae was attributed to chicken manure at the highest level (Table 5; ^{Wu and Zou, 2010}); however, the highest percentage of hyphae (48.56%) was observed with mushroom compost on 2 tha^-1. These results show that organic fertilizers cause a positive response in broad bean, particularly poultry manure. ^{López et al. (2001)} mentioned that organic fertilizers are a viable option to replace or complement inorganic fertilization.

Principal component analysis

Main components 1 and 2 explained 51.42% of the original total variation (Figure 1). These percentages are desirable to reliably interpret the approximate correlations observed in the bi-plot (^{Sánchez et al., 1995}; ^{Pérez et al., 2014}). In quadrant 1, there was positive correlation between yield (RG) with 100 seeds weight (P100S) and clean seeds (PSL), pods (NVP) and pod weight per plant (PVP), number and weight of stained seeds (NSM, PSM), greenness index (IV), plant height (AP) and floral knots (NNF). Treatments that interacted favorably with these variables were 3, 13 and 13 (RG 2.46, 2.63 and 2.23 t ha^-1). These facts indicate that the increase in yield is mainly due to direct and indirect improvements on them; thus, breeding, generation, validation, validation and technology transfer should benefit these interactions (^{Perez et al., 2014}).

Figure 1. Interrelationships between 40 treatments (in number) and 21 variables (in print). 1= San Pedro, without mycorrhiza and 1 ton of chicken manure; 2= San Pedro, without mycorrhiza and 2 tons of chicken manure; 3= San Pedro, without mycorrhiza and 3 tons of chicken manure, ..., 21 = Santiago, without mycorrhiza and 1 ton of chicken manure; 22= Santiago, without mycorrhiza and 2.0 tons of chicken manure; 23= Santiago, without mycorrhiza and 3 tons of chicken manure, ..., 40= Santiago, with mycorrhiza and 60N-60P-30K. 

These results are similar to those from ^{Neal and McVetty (1983)} who concluded that 68.5 to 76.4% of the variability in seed yield is due to more pods per plant (^{Chaieb et al., 2011}), seeds per pod (^{Alan and Geren, 2007}) and 100 seeds weight (^{Baginsky et al., 2013}). The colonization rate (% COL) was also correlated with root length (LRz), arbuscules percentages (% ARB), hyphae (% HIF), vesicles (% VES) and pink nodules (NDRs). This fact underlines that the application of mycorrhizae had a positive effect in both cultivars when chicken manure and mushroom is added at a dose of 3 t ha^-1 (T13, T19, T33 and T39).

In the above context, T13 (San Pedro with 3 t ha^-1 of chicken manure and G. fasciculatum) and T19 (San Pedro with 3 t ha^-1 mushroom compost and G. fasciculatum) significantly contributed to the improvement in floral knots (NNF, 15.8 and 16), plant height (AP, 121.33 and 123.43 cm), greenness index (IV, 52.90 and 52.66) and pods per plant (NVP, 25.93, 25.73); the first also did in weight of clean seed (PSL, 62.53 g) and 100 seeds weight (P100S, 301.83 g) and yield (RG 2.63 t ha^-1). T19 interacted positively with root length (LRz, 27.03), white nodules (NDBI, 12.5), colonization (COL, 72.2%) and arbuscules (ARB 71.46%) but these effects did not increase its RG. Treatment 6 (San Pedro with 3 t ha^-1 vermicompost) expressed the highest quality of clean seed. This fact is contradictory because in the presence of arbuscules there is higher efficiency nutrient absorption and symbiosis when there is available phosphorus, like chicken manure (^{Wu and Zou, 2010}).

In other studies it was observed that with G. fasciculatum improved growth of Vitis vinifera L. and phosphorus content was 15 times higher than in control (^{Alarcón et al., 2001}). ^{Robles et al. (2013)} reported an increase in root length of Agave angustifolia. The cv. Santiago with 3 t ha^-1 of chicken manure plus mycorrhizal improved pods weight per plant (PVP, 124.1), clean seeds (NSL, 32.63), weight of clean seeds (PSL, 62.4 g), grain yield (RG 2.4 t ha^-1) and arbuscules percentage (73.3) and hyphae (84.4), higher values than those reported by ^{Singh et al. (2013)}. These results are attributed to arbuscules are fungal structures haustorium type, that generate inside the cortical cells and contribute to increased capacity of absorption and utilization of nutrients by both participants in symbiosis (Alarcón and Ferrera, 2000). Hyphae provide increased absorption surface than root hairs and thus significantly increase the uptake of relatively immobile ions such as phosphate, copper and zinc. ^{Rojas and Ortuño (2007)} found a positive effect by inoculating mycorrhizal and adding chicken manure and vermicompost in vegetables (^{Manjarrez et al., 1999}).