The field experiment was conducted at the Experimental Site of Yunnan Agricultural University during October 2014 to May 2015. This site is dominated with red soil of pH 6.35. It contained organic matter of 14.4 g/kg, available K 116.4 g/kg, available P 24.2 g/kg, and alkali-hydrolyzed N 105.2 g/kg.

The fertilizers applied to the intercropped and monocropped wheat were a composition of urea, single superphosphate, and potassium sulfate. The application rate for wheat was 225 kg∙N/ha, 75 kg P₂O₅/ha, and 75 kg K₂O/ha. No organic fertilizer was applied. The application rates of N in the intercropped and monocropped faba bean were half of that in the wheat crops. The rates of P and K fertilizer in the faba bean were the same as that of wheat crops. The N, P, and K were applied once as a basal dressing on the faba bean. In the wheat field, 50% N fertilizer was applied as a basal dressing and the remainder was to pdressed twice, while P and K fertilizer were all applied once as a basal dressing.

The experiment included three treatments, i.e., monocropping of wheat, monocropping of faba bean, and wheat-faba bean intercropping. Each treatment was replicated four times. A total of 12 plots (each sized 4 m × 2 m) were arranged randomly. Wheat was sown by drilling with a row spacing of 0.2 m. Faba bean was sown by dibbling method with an inter-row spacing of 0.3 m and intra-row spacing of 0.2 m. We applied row-intercropping by planting six rows of wheat with 2 rows of faba bean. There were 14 rows in each intercropping plot, 19 rows in each plot of monocropped wheat, and 11 rows in each plot of monocropped faba bean.

Two to three soil samples were randomly collected in the monocropping treatment. In the intercropping treatment, soil samples were collected from rows where wheat and faba bean met. At harvest, the rhizosphere soil was collected by shaking soils apart from the root and then was well mixed. The soil samples required for the PLFA analysis were stored at −20˚C, and soil samples of AOA and AOB test were stored at −80˚C.

The whole plant was oven-dried at 150˚C for half-hour, then was oven-dried at 60˚C - 70˚C until constant weight, which referred to the biomass. The seeds of wheat or faba bean were collected from a 1 m² plot, air-dried to constant weight to represent the yield. Land equivalent ratio (LER) was used as an index of yield advantage which was calculated using the equation:

LER = ( Y i w / Y s w ) + ( Y i s / Y s s ) (1)

where Y_iw and Y_is represent the yield or biomass of wheat and faba bean on the total area of intercropping, respectively. Y_sw and Y_ss represent the yield or biomass of mono-wheat/faba bean on the whole area, respectively. LER value > 1 indicates an advantage from intercropping, and LER < 1 indicates a disadvantage from intercropping.

The soil microbial community was detected by the Phospholipid fatty acids (PLFA) analysis [18] [19]. Two grams of fresh soil samples were collected and the modified Bligh & Dyer method were applied for lipid extraction and PLFA analysis [20]. The soil samples were extracted with citric acid buffer (0.1 mol/L, pH 4.0), chloroform, and methanol in a volume ratio of 0.8:1:1.2. The extraction was acquired by silicic acid bonded solid-phase-extraction column (SPE-SI), eluted with chloroform, acetone and anhydrous methanol in sequence. The phospholipid fraction was dried with pure nitrogen gas and then hydrolyzed and saponified (methylated) with alkaline methanol to get phospholipid fatty acid methyl ester (FAME). The PLFA was determined according to the manufacturer’s instructions using the MIDI method, and C19 (methyl nonadecanoate) was used as an internal standard.

DNA was extracted from about 0.5 grams of fresh soil samples using the soil-specific DNA extraction kit, and the quality and quantity of DNA were tested using a NanoDrop spectrophotometer. The product of fluorescence quantitative PCR (qPCR) was determined by a fluorescence quantitative assay system. The SYBR Premix Ex Taq™ Perfect Real Time Kit (Dalian Bao Bioengineering Co.) was performed on a CFX96 Real-Time PCR System analyzer. The qPCR assays were carried out in a 20 μL reaction containing 2 μL DNA template, 10 μL SYBR Premix Ex Taq Perfect Real Time, 0.3 μL front primer, 0.3 μL rear primer, and 7.4 μL sterilized double distilled water. In the control group, sterile double distilled water was used as DNA template. The primers for ammonia-oxidizing bacteria and ammonia-oxidized archaea were ArchamoAF/Arch-amoAR [21] and Arch-amoAF/Arch-amoAR [22] , respectively. The primers and PCR conditions used in nitrifying bacteria are on Table 1.

Compared with monocropped wheat, the yield and biomass of intercropped

Note: M = A/C, R = A/G, W = A/T, S = G/C, Y = C/T, K = G/T, V = A/G/C, H = A/C/T, D = A/G/T, B = G/C/T, N = A/G/C/T.

wheat significantly increased by 12.6% and 15.8%, respectively (p < 0.05) (Table 2). But no significant effect of intercropping on the yield and biomass was found for faba bean. The LER values of grain yield and biomass in the intercropping system were all greater than one, indicating an obvious intercropping advantage.

A total of 37 PLFA were identified in the rhizosphere at harvest after wheat-faba bean intercropping, including 31 bacterial PLFA, 3 fungal PLFA, and 3 actinomyce PLFA (Table 3). The results showed that the bacterial PLFA and fungal PLFA in the rhizosphere of intercropped faba bean and actinomyces of intercropped wheat were significantly (p < 0.05) different from those of monocrops at harvest stage.

Principal component analysis (PCA) can reduce dimension mathematically to target the dominant variables through linear transformation from multiple variables. The 37 PLFA identified in the rhizosphere of wheat and faba bean under different planting patterns at harvest were analyzed to reveal the changes of microbial community and the dominant responding types of fatty acids. The first m principal components with the corresponding eigenvalues greater than one were selected as the principal component. According to the Kaiser standard, PC1 and PC2 with eigenvalues greater than one can explain most of the information of the variables.

The results of PCA are shown in Figure 1(a). The first principal component (PC1) and the second principal component (PC2) accounted for 49.3% and 16.8% variation of all variables, respectively. The PC2 showed significant difference between monocropped and intercropped faba bean, but no differences were

Note: M and I indicate monoculture and intercropping; values with different small letters are significantly different between intercropping and monoculture at a significance level of 5%. The explanations of all abbreviations apply to other figures and tables.

found between monocropped and intercropped wheat. Our finding revealed that the types of PLFA has no differences between monocropped and intercropped wheat, however, there were differences in microbial community between monocropped and intercropped faba bean. The distribution of most PLFA along the PC1 was shown in Figure 1(b), with the cumulative contribution rate over 80%. The sum of the following PLFA was used as a measure of Gram-positive bacteria (G+): 14:0 iso, 15:1 iso G, 15:0 iso, 15:0 anteiso, 16:0 iso, 17:0 iso, 17:0 anteiso, 18:0 10-methyl, TBSA. The sum of the following PLFA was used as a measure of Gram-negative bacteria (G−): 12:00, 12:0 iso 3OH, 16:1 w5c, 17:0 cyclo, 17:00, 18:1 w7c 11-methyl, 17:0 iso 3OH. The sum of the following PLFA was used as a measure of aerobic bacteria: 15:1 iso G, 15:0 iso, 15:0 anteiso, 16:0 iso, 17:0 iso, 17:0 anteiso, and 17:00. The PLFA 17:0 cyclo was used as a measure of anaerobic bacteria and 18:1 w9c was used as a measure of the fungi. The PLFA 18:0 10-methyl and TBSA were used as a measure of actinomycete.

The ratio of fungal:bacterial PLFA in rhizosphere soil of mono-faba bean was higher than that in intercrops; there were no significant intercropping effects on the ratio of fungal: bacterial PLFA in the rhizosphere of wheat (Figure 2).

The gene copies of AOA and AOB in the rhizosphere at harvest after intercropping are shown in Figure 3. There were no significant differences in AOB in the rhizosphere of wheat between monocrops and intercrops at harvest. For the rhizosphere of faba bean, the AOA and AOB abundance of monocrops was greater than that of intercrops.

Land management practices, plant species and physiological status have a significant effect on soil microbial activity and community structure in the intercropping

system [23]. The intercropping advantage was attributed to crop interactions and the changes in crop rhizosphere microbial activity [23]. In the present study, bacterial and fungal biomass in the rhizosphere of faba bean were significantly different between intercropped and monocropped faba bean (p < 0.05). The same pattern was found in actinomycete biomass in the rhizosphere of wheat (p < 0.05). Previous study indicated that a certain component of root exudates could promote the accumulation of Gram-positive bacteria and AM fungi in soil, and could also increase the total microbial biomass [24]. Root exudates contain sugars, amino acids, vitamins, and other substances, all of which could provide the nutrients and energies for the survival and reproduction of rhizosphere microorganisms [25]. Furthermore, different crops in the intercropping system could release specific root exudates, which could alter the structure of rhizosphere microbial communities to form a suitable rhizosphere environment. This could improve the overall metabolic activities of soil microbes, increase soil microbial diversification and maintain a healthy development of soil [25]. Li et al. used the PLFA analysis to confirm that the rhizosphere microbial community structure of maize and legume could be changed by intercropping in acidic soil [9]. Other Studies also showed that when the soil microbial community structure was richer and the microbial diversity was higher, the resistance to pathogen synthesis capacity would be also stronger [15] [26].

The ratio of fungal: bacterial PLFA reflects the variation range of relative amounts of fungi and bacteria and their relative abundance [27] [28]. Bacteria are generally considered to be the dominant component in fertile soils. A high fungal: bacterial PLFA ratio means a stable soil ecosystem [29]. However, other studies suggested there was no correlation between fungal and bacterial biomass, land use, and land management practices [30]. In our study, the ratio of fungal: bacterial PLFA in monoculture faba bean was significantly higher than that of intercropping, but the wheat rhizosphere of monoculture and intercropping did not differ. This may be because the utilization of carbon sources in the rhizosphere microorganisms of the faba bean is greater in the intercropping systemthan that of the wheat [31]. Alternatively, the secretion of organic acids and the release of acid phosphatase in the leguminous rhizosphere in the faba bean and intercropping systemcan may have resulted in a derease in pH in the gramineous crop rhizosphere [9] [32] which is not conducive to the survival of certain microorganisms. The amount of bacteria PLFA in the rhizosphere of the wheat monoculture and intercropping system was higher than that of faba bean because the application of nitrogen fertilizer could promote the growth and reproduction of bacteria and actinomycetes [33]. The increase of the species and number of bacteria, fungi and actinomycetes contribute to the formation of microbial diversity in the rhizosphere and increase the disease resistance of plants in intercropping systems [33] [35].

Nitrification is an important process in global nitrogen cycle and may contribute to nitrogen losses from the agroecosystem to the environment and result in water eutrophication and harmful trace gas emissions [36]. Microorganisms involved in the nitrification process are mainly ammonia-oxidizing bacteria (AOB) or archaea (AOA) and nitrite-oxidizing bacteria (NOB) [36]. In this study, the abundance of AOB was higher than that of AOA with soil pH of 6.4. Soil pH is one of the major drivers affecting the distribution of AOA and AOB populations. It was found that AOA populations were more resilient to low pH environment than AOB [37] , suggesting AOB were more suitable for survival and reproduction than AOA under the soil conditions in this study. Some studies showed that AOB is the main driver of nitrification in several typical soil in China [38] [39]. The wheat-faba bean intercropping could significantly improve the abundance of AOA and AOB and also change the structure of soil microbes, but there are many reasons which can influence the structure of AOB and AOA in the rhizosphere of the intercropping system, so the reasons for the difference should not be understood as a whole.