Elevated CO2 and nitrogen addition have minimal influence on the rhizospheric effects of Bothriochloa ischaemum

The influence of elevated CO2 and nitrogen (N) addition on soil microbial communities and the rhizospheric effects of Bothriochloa ischaemum were investigated. A pot-cultivation experiment was conducted in climate-controlled chambers under two levels of CO2 (400 and 800 μmol mol−1) and three levels of N addition (0, 2.5, and 5 g N m−2 y−1). Soil samples (rhizospheric and bulk soil) were collected for the assessment of soil organic carbon (SOC), total N (TN), total phosphorus (TP), basal respiration (BR), and phospholipid fatty acids (PLFAs) 106 days after treatments were conducted. Elevated CO2 significantly increased total and fungal PLFAs in the rhizosphere when combined with N addition, and N addition significantly increased BR in the rhizosphere and total, bacterial, fungal, Gram-positive (G+), and Gram-negative (G−) PLFAs in both rhizospheric and bulk soil. BR and total, bacterial, G+, and G+/G− PLFAs were significantly higher in rhizospheric than bulk soil, but neither elevated CO2 nor N addition affected the positive rhizospheric effects on bacterial, G+, or G+/G− PLFAs. N addition had a greater effect on soil microbial communities than elevated CO2, and elevated CO2 and N addition had minor contributions to the changes in the magnitude of the rhizospheric effects in B. ischaemum.


Results
Soil TOC, TN, and TP contents. A two-way ANOVA showed that elevated CO 2 did not have significant effect on SOC or TN content in either the rhizosphere or bulk soil, and N addition did not have significant effect on SOC, TN, or TP content (Table 1). Elevated CO 2 and N addition did not have interactive effect on SOC, TN, or TP content. Elevated CO 2 significantly affected TP content in both the rhizosphere and bulk soil (P < 0.05). TP content tended to decrease in response to elevated CO 2 ( Table 2).
Microbial respiration. Elevated CO 2 did not have significant effect on basal respiration (BR) or substrate-induced respiration (SIR) in the rhizosphere or bulk soil (Table 1). N addition significantly increased  Table 1. P values for the effects of elevated CO 2 (C), N addition (N), and their interaction on chemical and microbial properties in the rhizosphere and bulk soil. Significant P values are highlighted in bold.
BR in the rhizosphere (P < 0.001). BR for the rhizosphere was 74.7 and 101.2% higher in N2 than N0 at ambient and elevated CO 2 , respectively ( Fig. 1). BR was significantly higher in the rhizosphere than the bulk soil in all six treatments, and this positive rhizospheric effect was significantly affected by N addition (P < 0.001; Table 3, Fig. 2a). SIR was similar in the rhizosphere and bulk soil, indicating no rhizospheric effect (Figs 1b and 2b).

CLPP analysis.
A two-way ANOVA indicated that neither elevated CO 2 nor N addition had significant effect on AWCD, H, or D (Table 1). Elevated CO 2 and N addition did not have interactive effect on the functional diversity of the soil microbial communities in either the rhizosphere or the bulk soil (Table 4).

PLFA analysis.
A two-way ANOVA indicated that elevated CO 2 had significant effect on total and fungal PLFAs in the rhizosphere (Table 1). Compared with AN2, EN2 significantly increased total PLFA in the rhizospheric soil, and fungal PLFA was significantly higher in EN1 than AN1 in the rhizospheric soil (Fig. 3). N addition significantly increased total, bacterial, fungal, G + , and G − PLFAs in both the rhizosphere and bulk soil (  Table 2. Soil organic carbon (SOC), total nitrogen (TN), and total phosphorus (TP) contents (g kg −1 ) in the rhizosphere and bulk soil in the treatments. Different letters within a column indicate significant differences between treatments. Figure 1. BR (a) and SIR (b) in the rhizosphere and bulk soil in the treatments. Different lowercase letters indicate significant differences between treatments in the rhizosphere. AN0, ambient CO 2 and no N added; AN1 ambient CO 2 and N supply at a rate of 2.5 g N m −2 y −1 ; AN2, ambient CO 2 and N supply at a rate of 5 g N m −2 y −1 ; EN0, elevated CO 2 and no N added; EN1, elevated CO 2 and N supply at a rate of 2.5 g N m −2 y −1 ; EN2, elevated CO 2 and N supply at a rate of 5 g N m −2 y −1 .  Table 3. P values for the effects of elevated CO 2 (C), N addition (N), and their interaction on rhizospheric effects. Significant P values are highlighted in bold. and PC2, accounted for 81.44 and 1.71% of the variance, respectively (Fig. 4). The PLFA patterns differed significantly between the rhizosphere and bulk soil along PC1. Elevated CO 2 and N addition did not have significant interactive effect on the composition of microbial PLFAs (Table 1). Total, bacterial, G + , and G + /G − PLFAs were significantly higher in the rhizosphere than bulk soil (Fig. 5). The positive rhizospheric effect (variables were higher in the rhizosphere than the bulk soil) for total PLFA was significantly increased only by elevated CO 2 ( Table 3). The positive rhizospheric effects for bacterial, G + , and G + /G − PLFAs were not affected by either elevated CO 2 or N addition.

Discussion
Characteristics of the soil microbial communities under elevated CO 2 and N addition. Soil respiration is an important part of the global C cycle and the largest component of C flux from terrestrial ecosystems to the atmosphere 43,44 . Elevated CO 2 and N deposition can have profound impacts on soil respiration 30 . A meta-analysis found that N addition significantly increased soil respiration by 7.84% in grasslands 45 . N addition in our study significantly increased BR in the rhizosphere. This result was consistent with those by Luo et al. 46 and Zhang et al. 47 , who found that N application significantly increased soil respiration, which they attributed to N-induced increases in plant growth, especially root biomass. We also found that N addition significantly increased root biomass (Table 5). Soil respiration is generally sensitive to elevated CO 2 48 . Baronti et al. 13 and Liu et al. 49 observed increased soil respiration under elevated CO 2 . A meta-analysis by De Graaff et al. 50 found that soil respiration increased by 17.1% under elevated CO 2 and suggested that these increases could be due to microbial responses from changes in substrate availability. Elevated CO 2 in our study did not have significant impact on BR in either the rhizosphere or bulk soil, suggesting that 106 days elevated CO 2 might not significantly increase substrate availability to the community, although elevated CO 2 significantly increased root biomass ( Table 5).
The response of the soil microbial community to elevated CO 2 and N deposition depends on many factors, such as plant species, soil temperature and water content, and especially nutrient availability 23,51 . Previous studies have reported that elevated CO 2 or N addition increased, decreased, or had no significant impact on community structure 7,15,16,26,28,52 . These contradictory results could be due to differences of microbial substrate availability under different scenarios of climate change. Elevated CO 2 usually indirectly affects microbial communities by altering root biomass and exudation. Elevated CO 2 for 106 days in our study significantly increased plant biomass, and elevated CO 2 and N addition had significant interactive effect on plant root biomass, so total and fungal PLFAs significantly increased in the rhizosphere in the treatments with both elevated CO 2 and N addition. N addition directly increased soil N content and may have indirectly improved nutrient availability by changing Rhizospheric effects for BR (a) and SIR (b) in the treatments. *Above each bar indicates significant difference between the rhizosphere and bulk soil. Different lowercase letters indicate significant differences in rhizospheric effects between treatments. AN0, ambient CO 2 and no N added; AN1 ambient CO 2 and N supply at a rate of 2.5 g N m −2 y −1 ; AN2, ambient CO 2 and N supply at a rate of 5 g N m −2 y −1 ; EN0, elevated CO 2 and no N added; EN1, elevated CO 2 and N supply at a rate of 2.5 g N m −2 y −1 ; EN2, elevated CO 2 and N supply at a rate of 5 g N m −2 y −1 .  Table 4. Average well color development (AWCD) and functional diversity (Shannon index (H), and Simpson index (D)) of the soil microbial communities in the treatments in the rhizosphere and bulk soil. , and G + /G − PLFA (g) in the rhizosphere and bulk soil in the treatments. Different lowercase letters indicate significant differences between treatments in the rhizosphere, and different uppercase letters indicate significant differences between treatments in the bulk soil. AN0, ambient CO 2 and no N added; AN1 ambient CO 2 and N supply at a rate of 2.5 g N m −2 y −1 ; AN2, ambient CO 2 and N supply at a rate of 5 g N m −2 y −1 ; EN0, elevated CO 2 and no N added; EN1, elevated CO 2 and N supply at a rate of 2.5 g N m −2 y −1 ; EN2, elevated CO 2 and N supply at a rate of 5 g N m −2 y −1 .
the conditions of plant growth ( Table 5). The effects of N enrichment on the communities are associated with soil critical N loads or N saturation theories 53 . These theories propose that the effect of N enrichment on ecosystem functions would switch from stimulation to inhibition when the ecosystem reaches a critical N loading or saturation level 23,54,55 . The N-saturation theories suggest that low N additions usually increase soil microbial biomass and microbial diversity and that high N addition would decrease them 25,26 . Plant growth on the Loess Plateau is restricted by soil N content, and N addition in our study significantly increased soil microbial biomass and shifted the community structure.
Rhizospheric effects under elevated CO 2 and N addition. Rhizospheres are zones of higher microbial turnover and activity, because they are adjacent to plant roots 56 . Many important aspects of plant-soil interactions are mediated by rhizospheric processes, including nutrient acquisition and root colonization by rhizospheric microorganisms. Microbial activity is higher and more diverse in rhizospheres than bulk soil 33 . We also found that BR and total, bacterial, G + , and G + /G − PLFAs were significantly higher in the rhizosphere than bulk soil. The higher G + /G − PLFAs in the rhizosphere than the bulk soil indicated that the rhizosphere communities were more heterotrophic via increases in C inputs when the plants were exposed to elevated CO 2 and N addition 57,58 . Rhizospheric effects can be affected by elevated CO 2 and N addition 36,59 . The types and amounts of organic root exudates can be altered when plants are exposed to high levels of CO 2 , and these changes may affect rhizospheric microbial activity and community composition. Lee et al. 4 reported that rhizospheric microbes responded to elevated CO 2 more strongly than the microbes in bulk soil. Our results showed that elevated CO 2 significantly increased the rhizospheric effects of total PLFA, as expected, because rhizospheric microbial communities are sensitive to elevated CO 2 . The rhizospheric effects of other microbial variables (such as bacterial PLFA, G + PLFA, and G − PLFA), however, responded weakly to elevated CO 2 . The impact of N addition on rhizospheric effects can be affected by many factors, such as plant species, soil type, soil chemical properties, and the amount and duration of N addition 35 . Phillips and Fahey 60 found that N fertilization had a positive, negative, or no impact on rhizospheric effects for trees, depending on the tree species and soil variables. Ai et al. 61 reported that long-term inorganic N addition reduced rhizospheric effects in a wheat-maize rotation system. N addition in our study did not significantly affect the rhizospheric effects for the soil variables, except BR, consistent with the results by Zhu et al. 36 , who also reported that N fertilization had minimal influence on the rhizospheric effects of two grass species. N is the limiting factor for plant growth on the Loess Plateau, and N addition in our study significantly increased plant total biomass and root biomass, so the root-derived C inputs to the soil (e.g. root exudates) may have been significantly affected by the N addition, and the available substrates in the rhizosphere (e.g. NH 4 + , NO 3 − ) may also have been significantly absorbed by the plants (Table 5), which may account for the lack of significant shifts in the rhizospheric effects 34,62 .
Evaluating the effects of elevated CO 2 and N deposition on soil microorganisms in the rhizosphere and bulk soil is challenging because of their high diversity. Our PLFA analysis found that elevated CO 2 and N addition had significant effects on the soil microbial community, especially in the rhizospheric soil. Both elevated CO 2 and N addition contributed little to the changes in the magnitude of the rhizospheric effects, perhaps due to the low resolution of PLFAs for classifying soil microbial communities. As science and technology have developed, especially in the last decade, high-throughput molecular technologies have been developed for characterizing microbial communities, including high-throughput DNA/RNA sequencing, PhyloChio, GeoChip, mass spectrometry-based proteomics for community analysis, and metabolite analysis 63 . In future studies, these molecular methods maybe able to provide a more comprehensive understanding of microbial responses to scenarios of global climate change. soil; AN0, ambient CO 2 and no N added; AN1 ambient CO 2 and N supply at a rate of 2.5 g N m −2 y −1 ; AN2, ambient CO 2 and N supply at a rate of 5 g N m −2 y −1 ; EN0, elevated CO 2 and no N added; EN1, elevated CO 2 and N supply at a rate of 2.5 g N m −2 y −1 ; EN2, elevated CO 2 and N supply at a rate of 5 g N m −2 y −1 .

Conclusions
N addition significantly increased BR in the rhizosphere and increased total, bacterial, fungal, G + , and G − PLFAs in both the rhizosphere and bulk soil, but elevated CO 2 only significantly increased total and fungal PLFAs in the  (f), and G + /G − PLFA (g) in the treatments. * Above each bar indicates significant difference between the rhizosphere and bulk soil. Different lowercase letters indicate significant differences in rhizospheric effects between treatments. AN0, ambient CO 2 and no N added; AN1 ambient CO 2 and N supply at a rate of 2.5 g N m −2 y −1 ; AN2, ambient CO 2 and N supply at a rate of 5 g N m −2 y −1 ; EN0, elevated CO 2 and no N added; EN1, elevated CO 2 and N supply at a rate of 2.5 g N m −2 y −1 ; EN2, elevated CO 2 and N supply at a rate of 5 g N m −2 y −1 .
SCIEntIFIC REpORtS | 7: 6527 | DOI:10.1038/s41598-017-06994-3 rhizosphere when combined with N addition. These results demonstrated that N addition had a larger impact on the soil microbial communities than elevated CO 2 . Contrary to our second hypothesis, the rhizospheric effects of soil microbial variables were not significantly affected by elevated CO 2 and N addition. These results suggest that the rhizospheres of B. ischaemum exert a more important control of community composition and structure than short-term elevated CO 2 and N addition. Experimental design and sample collection. Each plastic pot (20 cm high, 15 cm in diameter) was separated vertically into two concentric zones, a central root zone and a root-free zone, by 25-μm nylon mesh bags (20 cm high, 9 cm in diameter) buried in the centers of the pots, enabling the passage of water and nutrients but not roots. The seeds of B. ischaemum were sown in the mesh bags on 1 June 2014. The soil-water content was maintained above 80% FC during the entire experiment.

Materials and Methods
The experiment began on 1 August 2014 after the seedlings were thinned to three per pot. The pots were transferred to two closed climate-controlled chambers (AGC-D001P, Qiushi Corp., Hangzhou, China) programmed at 13 h of light (28 °C, relative humidity (RH) of 50%, 300 μ (photons) m −2 s −1 ) from 7:30 to 20:30 and 11 h of dark (22 °C, RH of 55%). The CO 2 concentrations in the two chambers were maintained at 400 (ambient) and 800 (elevated) μmol mol −1 until the end of the experiment. An automatic control system was used to adjust the CO 2 to the desired concentration in each chamber by regulating the influx rate of pure CO 2 to the air blower. Each chamber housed three N-addition treatments (0 (control), 2.5, and 5 g N m −2 y −1 ). The N-addition treatments began on 18 August 2014. For each pot, NH 4 NO 3 was dissolved in deionized water and then added to the pot soil, except for the control treatment where an equal volume of deionized water was added. N was added a total of six times during the experiment, with a frequency of every 15 days. The 0, 2.5, and 5 g N m −2 y −1 treatments received 0, 0.021, and 0.042 g NH 4 NO 3 , respectively, each time. All pots were weighed daily at 18:00, and water was added via plastic pipes to maintain soil-water contents above 80% FC. A total of six treatments were thus tested: ambient CO 2 but no N added (AN0), ambient CO 2 and 2.5 g N m −2 y −1 (AN1), ambient CO 2 and 5 g N m −2 y −1 (AN2), elevated CO 2 but no N added (EN0), elevated CO 2 and 2.5 g N m −2 y −1 (EN1), and elevated CO 2 and 5 g N m −2 y −1 (EN2). Each treatment had five replicates.
The experiment was completed on 15 November 2014. The soils of the root zone (adhering to the roots in the mesh bag) and the root-free zone (>1.5 cm outside the mesh bag) were collected and sieved through a 2-mm mesh. One subsample of each type of soil was air-dried, crushed, and passed through a 0.25-mm mesh for the determination of chemical properties, another subsample was stored at 4 °C for respiration and Biolog analysis, and a third subsample was stored at −20 °C for PLFA analysis.
Soil chemical properties. SOC content was determined by wet digestion with a mixture of potassium dichromate and concentrated sulfuric acid, TN content was determined by the semimicro Kjeldahl method after digestion by H 2 SO 4 , and TP content was determined colorimetrically after wet digestion with H 2 SO 4 + HClO 4 .  Table 5. Effects of elevated CO 2 concentration (C) and N addition (N) on plant biomass and nitrogen content. Different letters within a column indicate significant differences (P < 0.05) based on Duncan's multiple range test. Significant P values are highlighted in bold. AN0, ambient CO 2 and no N added; AN1 ambient CO 2 and N supply at a rate of 2.5 g N m −2 y −1 ; AN2 , ambient CO 2 and N supply at a rate of 5 g N m −2 y −1 ; EN0 , elevated CO 2 and no N added; EN1, elevated CO 2 and N supply at a rate of 2.5 g N m −2 y −1 ; EN2 , elevated CO 2 and N supply at a rate of 5 g N m −2 y −1 .
The basal respiration (BR) was determined by placing 10 g of each soil sample, moistened to 50-60% of the field capacity, in a hermetically sealed flask equipped with a rubber septum for gas sampling. The samples were then incubated at 28 °C for 7days under aerobic conditions, and the CO 2 released was measured 0.5, 1, 2, 3, 4, 5, 6, 7 days after incubation with an infrared gas analyzer (QGS-08B, Befen-Ruili Analytical Instrument Co. Ltd., Beijing, China). Soil samples were amended with 0.6% (w/w) glucose before incubation for determining substrate-induced respiration (SIR). The BR and SIR data are expressed as μg CO 2 g −1 dw h −1 .

Community-level physiological profile (CLPP
where x i is the optical density measured at 590 nm for substrate in the EcoPlates, c is the OD of the control well, 31 is the number of C sources, p i is the ratio of the absorbance in each well to the sum of absorbance for all wells, and n is the total number of C sources. PLFAs. The method for phospholipid extraction was adapted from Buyer et al. 67 . Briefly, 3 g of lyophilized soil were placed in a 30-ml centrifuge tube with a Teflon-lined screw cap. The fatty acids were directly extracted from the soil twice by adding 3.6 ml of citrate buffer (pH 4.0), 4 ml of chloroform, and 8 ml of methanol. The PLFAs were separated from neutral and glycolipid fatty acids by solid-phase-extraction chromatography. After mild alkaline methanolysis, the PLFA samples were qualitatively and quantitatively analyzed using an Agilent 7890 gas chromatograph (Agilent Technologies, Santa Clara, USA) equipped with an autosampler, split-splitless injector, and flame ionization detector. The system was controlled with Agilent ChemiStation and MIDI Sherlock software (Microbial ID, Inc., Newark, USA). An external standard of 19:0 methyl ester was used for quantification.
We selected the following PLFA signatures to serve as indicators of specific microbial groups: iso-and anteiso-branched fatty acids for Gram-positive (G + ) bacteria 68 , monounsaturated and cyclopropyl 17:0 and 19:0 fatty acids for Gram-negative (G − ) bacteria 69 , and 18:2w6c for fungi 70 . Total biomass was obtained by summing the concentrations of all fatty acids detected in each soil sample. Statistical analysis. All results are expressed as means ± standard deviations. Rhizospheric effects were calculated as the percent difference between the rhizospheric and bulk-soil samples for each measured variable 34,36 . Student's t-tests were used to compare the values between the rhizosphere and bulk soil to indicate the statistical significance of the calculated rhizospheric effect. SOC, TN, and TP contents and microbial functional diversity did not differ significantly between the rhizosphere and bulk soil, so we have not reported the rhizospheric effects for these variables. Two-way analyses of variance (ANOVAs) at a probability level of 0.05 were used to assess the effects of elevated CO 2 , N addition, and their interaction on the biochemical properties of the rhizospheric and bulk soil and the rhizospheric effect for each variable. Means were compared using Duncan's multiple range test for significant differences (P < 0.05). A principal component analysis examined the PLFA community structure among the treatments. The above statistical analyses were performed using SPSS 16.0 (SPSS Inc., Chicago, USA).