Rhizospheric bacteria were isolated and purified from the rhizosphere soil of wild C. sachalinensis; 37 predominant strains were screened and cryopreserved. The sampling site was a mountainous area in Lianshanguan town, Benxi City, Liaoning Province, China (41°24’ N, 124°17’ E), located at an altitude of 512-546 m asl, with a land slope of 20° to 45°. The climate in this area is classified as humid in the northern temperate zone. Annual precipitation is approximately 800-900 mm and average annual temperature is 6.1°C to 7.8°C. The length of the frost-free season is approximately 156–172 d. The area has predominantly Brunisolic soil.

Seed germination experiments were conducted at a constant temperature of 20°C in an incubator (LHS-150HC-I; Shanghaiyiheng, China). These experiments were used for the pre-screening of growth-promoting and growth-inhibiting rhizospheric bacteria. For the preliminary screening of bacteria, seeds were treated with the suspension of the 37 bacterial strains previously isolated from the C. sachalinensis rhizosphere soil. Uniform-sized plump, stratified and fresh seeds of C. sachalinensis were used for the preliminary selection of efficient root-colonizing bacterial strains. Shelled C. sachalinensis seeds were disinfected with 0.5% potassium permanganate solution and placed into sterilized Petri dishes containing a filter paper. Twenty C. sachalinensis seeds and 6 mL bacterial suspension (prepared as described in below) that was diluted five times were placed into each Petri dish and cultivated at a constant temperature (20°C). This experiment was replicated three times. Next, 1 mL sterile water was added to each Petri dish, including the control treatment (which did not include the bacterial suspension), every alternate day to ensure seed germination. Germination percentage and seedling growth were determined after 7 d. Strains with evident germination promoting or inhibiting activities were selected.

The C. sachalinensis seeds were sown on April 25 in a shelter house with an open roof at the Shenyang Agricultural University (41°83’ N, 123°56’ E). The elevation, averages of accumulated temperature, annual sunshine hours, annual frost-free period and mean annual precipitation were 76.2 m, 3281°C, 2372 h, 146–163 d and 721 mm, respectively. The growth conditions were not controlled. Seedlings were planted in 12 cm × 13 cm feeding blocks and grown for 3 months in the culture substrate consisting of garden soil with 20% turfy soil. The basic physical and chemical properties of the substrate were as follows: pH, 7.18; soil organic matter, 76.61 g/kg; alkaline-hydrolyzable nitrogen, 296.3 mg/kg; available phosphorus, 35.51 mg/kg; and potassium, 32.17 mg/kg. Uniform-sized seedlings were selected for treatment with bacterial suspension (50 mL/pot) and 10 replicate seedlings were used for each treatment. The mean plant height was 24.5 ± 2.8 cm before treatment. Seedlings treated with an equal amount of sterile water were used as the control. Inoculations were performed every 3 d for 30 d. Morphological changes in seedling growth and development and related indices of respiratory metabolism and rhizosphere environment were determined at the end of the 30-d treatment.

Microbial counts. A dilution plate count method was used for estimating the number of microorganisms (Johnson & Curl, 1972). Bacteria, fungi and actinomycetes were counted on peptone beef, Martin Substratum and Gause 1 cultural media, respectively. All materials for the cultural medium were purchased from Sigma Chemical Company (St Louis, MO, USA). The numbers of colonies per plate were determined by diluting the samples in deionized water to 10^–4, 10^–5 and 10^–6 g/mL, and 0.1 mL soil solution was added to each plate. The plates were incubated at 28°C for 1 d for bacteria, 3 d for fungi and 7 d for actinomycetes, and then the developing colonies were counted. Three plates were used for each treatment.

Acquisition of efficient root-colonizing bacteria.Efficient colonizers are defined as those strains that yield >5 colonies at the highest dilutions. In this experiment, soil was diluted to 1:1,000,000 with three replicates. Bacterial strains were cultured in peptone beef medium for 2–3 d, and a single developing colony was purified by culturing in peptone beef for 2–3 times. The purified bacteria were stored at 4°C.

Preparation of bacterial suspension. The obtained bacteria were cultured in peptone beef medium at 28°C for 24 h and then inoculated in liquid medium (peptone beef medium without agar) with shaking. The supernatant was collected after centrifugation and colorimetric analysis was performed at 600 nm by using a spectrophotometer (Agilent HP 8453; Agilent Technologies, Waldbronn, Germany). The OD₆₀₀ was adjusted to 0.4 with sterilized water and this bacterial suspension contained approximately 10⁸ CFU/mL (Peng et al., 2007).

Identification of bacterial strains by using 16S rDNA gene sequence analysis.Total DNA was extracted as previously described (Laguerre et al., 1992). Universal primers were used for polymerase chain reaction (PCR) amplification (forward primer, 8-27F: 5’-AGAGTTTGATCMTGGCTCAG-3’; reverse primer, 1378-1401R: 5’-CGGTGTGTACAAGGCCCGGGAACG-3’). PCR amplification was performed in a 50 μL reaction volume: 5 μL Taq DNA polymerase, 2.5 mM MgCl₂, 0.2 mM of each dNTP, 0.4 μM each of forward and reverse primers, 40 ng template DNA, 5 U of Taq DNA polymerase and 32 μL ddH₂O. The PCR conditions were as follows: initial denaturation at 95°C for 5 min, followed by 30 cycles of denaturation (1 min at 95°C), annealing (30 s at 60°C) and extension (1.5 min at 72°C) and a final extension at 72°C for 10 min. Amplified DNA was examined by gel electrophoresis on a 1.0% agarose gel. PCR products were purified using the PCR extraction kit (DNA Fragment Purification Kit vers. 2.0; Takara) and sent to Takara Biotechnology Co. Ltd., Dalian, for sequencing. The sequencing results were blasted to GenBank (http://www.ncbi.nlm.nih.Gov) for sequence alignment in order to obtain the closest sequence homology. Criterion for identification at the genus level was defined as a 16S rDNA sequence similarity of ≥97% with that of the prototype strain sequence in GenBank.

Root viability.The triphenyl tetrazolium chloride (TTC) method (Lindström & Nyström, 1987) was used to determine root viability. TTC is reduced mainly by dehydrogenases, which are associated with mitochondrial function in plants (Bagniewska-Zadworna, 2008). Higher root viabilities suggest higher physiological activities of dehydrogenases in the roots. Briefly, 500 mg (diameter <2 mm) of fresh fine roots were placed into beakers containing 5 mL 0.4% TTC solution and 5 mL Na₂HPO₄-KH₂PO₄ (pH 7.0) and incubated for 2 h at 37°C in the dark. Subsequently, triphenylformazan was extracted with ethyl acetate and the absorbance was measured at 485 nm. Root viability was calculated as C/(1,000·W·h) [mg TTF/(g·h)], where C represents reduction of tetrazolium, W represents root weight and h is time (hours).

Total root respiration rates.Root respiration rate was measured as oxygen consumption by using an Oxytherm oxygen electrode (Hansatech, England) according to a modified method by Bouma et al. (2001). The total respiration rate was determined by slicing new root tissue samples (1.5 mm wide, 2–3-cm long, 50 mg) into 2-mm pieces and setting aside for 15 min to eliminate the influence of wound respiration. Next, 0.05 g of each sample was weighed for measurement and six replicates were incubated at 25°C ± 1°C. Total respiration rates were determined as root O₂ uptake per unit of fresh weight (μmol O₂/min/g FW) in the assay buffer without inhibitors.

Root respiratory pathways.Root respiratory pathways were measured according to Yu & Pan (1996). The capacities of EMP, tricarboxylic acid cycle (TCA) and PPP pathways were determined using 0.5 M NaF, malonic acid and Na₃PO₄ as inhibitors, respectively. The COX and AOX pathways were inhibited using 0.1 M NaCN and salicylhydroxamic acid, respectively. Phosphate buffer (0.2 M, pH 6.8) was used as the reaction medium. Each respiratory pathway capacity was determined three times. The percentage of each respiratory pathway was calculated as [(Total respiration rate–Residual respiration rate)/Total respiration rate]. Residual respiration rates represented respiration rates after the addition of the corresponding inhibitor.

Key root respiratory enzymes.Fresh root tissues (0.5 g) were pulverized in 3 mL extraction buffer (100 mM Tris-HC1, pH 7.5) by using a mortar and pestle chilled in ice. The mixture was centrifuged for 30 min at 10,000 g at 4°C and the supernatant was assayed immediately for 3 min in 3.0 mL reaction mixtures. The activities of hexokinase (HK), pyruvate kinase (PK) and phosphofructokinase (PFK) were measured using the methods of Samuelov et al. (1991), Sridhar et al. (2000) and Ling et al. (1966), respectively. The malate dehydrogenase (MDH) and succinodehydrogenase (SDH) activities were measured according to Ouyang (1985). The isocitrate dehydrogenase (IDH) activity was measured according to Plaut (1969). The glucose-6-phosphate dehydrogenase (G-6-PDH) and 6-phosphate glucose dehydrogenase (6-GPDH) activities were measured as described by Lamed & Zeikus (1980).

Root respiratory substrates and intermediate products.Concentrations of sucrose and starch, pyruvic acid and citric acid were determined using anthrone colorimetry (Hodge, 1962), 2,4-dinitrophenylhydrazine (Katsuki et al., 1961) and enzymatic reaction colorimetry, respectively.

Biomass and soil properties. Seed germination and seedling growth were determined using a ruler and balance. The concentration of soil organic matter, alkali-hydrolyzable N, available P and available K were measured according to Shi et al. (1996).

Figures were drawn in Microsoft Excel 2003. Statistical analyses were performed using Statgraphics (STN). The response variables considered were: seed germination percentage; seedling height; fresh weight of seeds; root viability; root respiratory rates; percentage of each root biochemical respiratory pathway; key enzyme activities of EMP pathway, TCA pathway and PPP pathway; percentage of each root electron transport respiratory pathway; concentration of sucrose and starch; concentration of pyruvate and citrate of C. sachalinensis seedlings; increases of seedling height and seedling leaves; fresh weights of shoots and roots; dry matter weights of shoots and roots; number of bacteria, fungi and actinomycetes; percentage of treated bacteria in the total bacteria; alkali-hydrolyzable N concentration; and available P and K and organic matter concentrations. As explanatory variable for each analysis of variance (ANOVA) we used the bacterial strains. According to normality, the data fulfilled the presumptions of ANOVA. Differences between means were considered significant when the p value of the ANOVA F test was <0.05. All p values obtained from multiple comparisons were corrected using Tukey’s highly significant difference (HSD) method. p values of the one-way ANOVAs of strains (B) are indicated. *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.

Of the 37 bacterial strains, 31 were found to influence seed germination, seedling height and fresh weight either positively or negatively compared to the control treatment (Fig. 1). Seed germination rate, seedling height and fresh weight were significantly different between the control and cultures treated with Str. 6, 9, 12, 14, 30, 34, 35 and 37. Seed germination rate and seedling height and weight were significantly higher in plants cultured with Str. 12, 34, 35 and 37 than those of the control, indicating that these strains facilitated plant growth. In contrast, these indices were significantly lower in plants cultured with strains 6, 9, 14 and 30 than those of the control, indicating that these strains inhibited plant growth.

The taxonomic status of the efficient root-colonizing bacteria was identified and further experiments were performed to determine their effects on C. sachalinensis seedlings. The 16S rDNA gene sequence analysis revealed that strains 34, 35 and 37 were Bacillus cereus (GenBank Acc. No. HM245778.1), Staphylococcus sp. (AY940424.1) and Pseudomonas fluorescens (HM008949.1), respectively. Strain 12 was an unknown bacterium. Growth-inhibiting Str. 6, 9, 14 and 30 were identified as Stenotrophomonas maltophilia strain YLZZ-2 (EU022689.1), B. cereus (HM003208.1), B. cereus (HM055982) and P. fluorescens (DQ095904.1), respectively.

Root viability and respiration rate were significantly different among the eight dominant bacterial treatments and the control treatment (Fig. 2a). The viability of roots inoculated with growth-promoting Str. 12, 34, 35 and 37 was significantly higher than that of the control, with plants treated with growth-promoting Str. 37 showing the highest viability (42% higher than that of the control). The respiration rate of the control roots and those inoculated with growth-promoting Str. 12, 34, 35 and 37 were not significantly different (Fig. 2b). In contrast, the viability and respiration rates of roots inoculated with growth-inhibiting Str. 6, 9, 14 and 30 were significantly lower than those of the control, with plants treated with strain 6 showing the lowest viability (57.8% lower than that of the control) and respiration rate (59.5% lower than that of the control; Fig. 2b).

EMP and TCA were the dominant respiratory pathways in the roots of C. sachalinensis seedlings grown under normal conditions. TCA accounted for approximately 60% of the total respiratory rate, whereas PPP accounted for 24% (Fig. 3a). TCA, the major respiratory pathway in the control treatment, was dominant in the roots inoculated with the growth-promoting Str. 12 and 35. EMP was enhanced and TCA was reduced by the growth-promoting Str. 34. Roots inoculated with growth-promoting Str. 37 showed an increase in the proportion of TCA and reduction in PPP compared to those of the control. The activity of the TCA pathway was remarkably reduced in the roots inoculated with growth-inhibiting Str. 6, 9, 14 and 30. The proportion of EMP increased, but that of TCA declined significantly in the roots inoculated with growth-inhibiting Str. 9 and 14 compared to those of the control, indicating that these two bacterial strains inhibited normal root respiration in C. sachalinensis. In addition, the proportion of EMP remained unchanged in plants inoculated with Str. 6 and 30 compared to that in the control. Further, the EMP and PPP pathways were the dominant pathways that were affected by bacterial inoculations.

The proportion and direction of the basic biochemical pathways are regulated by various enzymatic activities. HK, PFK and PK are the key enzymes that regulate the EMP pathway. The key enzymatic activities in EMP were unchanged by the growth-promoting Str. 12, 34, 35 and 37 (Fig. 3b). In contrast, the activity of the key EMP enzymes changed significantly in the roots inoculated with the growth-inhibiting Str. 6, 9, 14 and 30. The activity of HK changed most significantly after inoculation with the growth-inhibiting bacterial strain 14, increasing by 4.4 times relative to that of the control. The activities of PFK and PK decreased significantly in the roots inoculated with Str. 6 and 30, but their activities were increased significantly by Str. 9 and 14.

TCA is mainly regulated by the activities of SDH, MDH and IDH. IDH activity was increased significantly by growth-promoting Str. 12, 34 and 35 and was decreased significantly by growth-inhibiting Str. 6, 9, 14 and 30. MDH activity was unchanged with the exception of a significant decrease in the presence of Str. 9 and a significant increase in the presence of Str. 34. SDH activity increased significantly in the presence of Str. 12, 35 and 37 and Str. 9 and 30 and was 3.9 times higher in the roots inoculated with Str. 37 than that in the control (Fig. 3c).

G-6-PDH and 6-GPDH are the key enzymes in the PPP pathway and G-6-PDH is the dominant enzyme. The activities of these enzymes decreased in the roots inoculated with the Str. 12, 34, 35 and 37, especially their activities decreased by 40.9% and 55.2%, respectively, in the presence of Str. 37 (Fig. 3d). The activity of G-6-PDH was increased by Str. 6, 9, 14 and 30 and especially by Str. 6 (82% increase). The activity of 6-GPDH was increased by Str. 6 and 9.

Under normal growth conditions, COX is the primary pathway of the respiratory electron transport chain (control; Fig. 3e). The proportion of COX was increased by Str. 34 and 37 and that of AOX was increased by Str. 35. The proportion of COX decreased significantly and that of AOX increased significantly in the roots inoculated with Str. 6, 9, 14 and 30.

Sucrose and starch are the primary substrates for respiratory metabolism under normal conditions. Sucrose concentration increased in the roots inoculated with Str. 35 and 37, but decreased in the roots treated with Str. 6, 9, 14 and 30 (Fig. 3f). Starch concentrations were significantly higher in the roots exposed to Str. 12 and 35 than those in the control, and they were significantly lower in the roots exposed to Str. 6, 14 and 30.

Pyruvic acid and citric acid are the key products of EMP and TCA. The pyruvic acid concentration decreased significantly in all the treatments, except for those treated with Str. 9 and 14. Citric acid concentration increased significantly in the roots exposed to the growth-promoting bacterial strain 37 and decreased significantly in the roots exposed to Str. 6, 9, 14 and 30. Other treatments did not affect citric acid concentration (Fig. 3g).

Compared with those of the control, the height, leaf number and shoot and root dry weights changed significantly in the seedlings inoculated with the eight strains of rhizobacteria (Table 1). All these variables increased in the seedlings whose roots were inoculated with Str. 12, 34, 35 and 37; the height and leaf number increased by 77.7% and 86.6%, respectively, in the C. sachalinensis seedlings treated with Str. 37. These variables decreased in the seedlings treated with Str. 6, 9, 14 and 30. The height and leaf number decreased by 58.7% and 63.6%, respectively, in the C. sachalinensis seedlings treated with Str. 6.

In the present study, the number of fungi significantly decreased, whereas the content of alkali-hydrolyzable N and available P increased in the rhizosphere soil after treatment with Str. 12, 34, 35 and 37. In contrast, after treatment with Str. 6, 9, 14 and 30, the fungal populations significantly increased, whereas the content of alkali-hydrolyzable N and available P varied slightly (Tables 2 and 3).