Screening plant growth-promoting bacteria from the rhizosphere of invasive weed Ageratina adenophora for crop growth

Site description and sampling strategy

Three sites where A. adenophora invasion had occurred in different floristic regions (Li et al., 2015) of Yunnan Province, China, were chosen for study (Fig. 1). Their floristic, climatic, and soil characteristics are listed in Table 1. The site at Kunming (ca. 50 × 50 m²) was located 50 m from a conifer-broadleaf forest. The site at Yuanjiang (ca. 50 × 30 m²) was located 30 m away from a conifer-broadleaf forest, where A. adenophora grew sporadically among a range of native plant species. The site at Yunlong (ca. 60 × 30 m²) was located in a valley where A. adenophora grew in large patches separated by several different native shrubs and grasses.

Three soil samples were collected from the rhizospheres of A. adenophora at each study site. Each individual sample site was separated from its neighbor by at least 2 m and all were located at the same altitude. To collect rhizosphere soil samples, individual plants were pulled up and shaken to remove the soil loosely attached to the roots, leaving only firmly attached soil, which was then used in the experiments described here. Individual soil samples were placed into sterile Whirl-Pak sample bags (Nasco, Fort Atkinson, WI, USA) and transported to the laboratory within 5 h of collection. Approximately 500 g of each soil sample was homogenized (medium treatment for 1 min) in a Waring heavy-duty blender (Lab-Biogen, Kunming, China), which was carefully cleaned after each use with 70% ethanol and passed through a 2 mm sieve before isolation of bacterial strains exhibiting PGP traits. Individual soil samples were collected from each study site on the same day between 15 August and 5 September, 2019.

Isolation and characterization of bacterial strains with in vitro PGP abilities

A total of 10 g soil from each soil sample was shaken (200 rpm) in a 500 mL flask with 90 mL sterile distilled water and glass beads for 10 min to isolate and characterize bacterial strains with in vitro PGP abilities (PGP bacteria). The flask was left at room temperature for 30 s before the suspensions were serially diluted in a 10× dilution series with sterile distilled water. The suspensions were then used to isolate bacterial strains on agar plates containing King’s B, Luria-Bertani (LB), Lowenstein-Jensen, R2A, and Gauze’s synthetic medium no.1 agar media to maximize the isolation of different PGPR strains. A total of 400 colonies from the three soil samples from each study site, based on different colony morphologies, were subcultured onto LB medium and screened for their abilities to fix nitrogen and produce IAA and ACC deaminase, as described below.

Identification of PGP bacteria by 16S rRNA sequencing

Genomic DNA was isolated from bacterial isolates using the Lysis Buffer for Microorganism to Direct PCR (Takara, Shanghai, China) according the protocol provided. 16S rRNA genes were amplified using universal primers 27f (5′-AGA GTT TGA TCM TGG CTC AG-3′) and 1492r (5′-CGG TTA CCT TGT TAC GAC TT-3′) (Lane, 1991) with 2×Taq MasterMix (Sangon Biotech, Shanghai, China). 16S rRNA gene amplicons were sequenced from both ends with an Applied Biosystems 3710X1 DNA Sequencer (Thermo Fisher Scientific, Shanghai, China). The nucleotide sequences obtained were aligned into consensus sequences using Mega11 software (Kumar, Tamura & Nei, 1993) and were then analyzed using BLASTn (NCBI) to identify their nearest neighbors. Distance trees showing all PGP bacterial strains were built with the Mega11 software.

Chamber germination experiments

The three PGP bacterial strains with the greatest abilities for N fixation, IAA and ACC deaminase production, respectively (KM_A34, KM_C04, and KM_A57) (Table S1) were selected and their PGP effects on the germination of soybean (Giycine max L. Merrill), faba bean (Vicia faba L.), wheat (Triticum aestivum L.), pea (Pisum sativum L.), Chinese cabbage (Brassica campestris L.), and A. adenophora were examined. Seeds of these crop plants were purchased from a local seed market. A. adenophora seeds were collected from A. adenophora plants from the Kunming site in May 2019.

Chamber experiments were designed with complete randomization, with six replicates per treatment. They included a set of non-inoculated controls set up according to the procedures described by Amogou et al. (2018) with slight modifications. Briefly, seeds were disinfected by soaking for 2 min in a sodium hypochlorite solution (0.024%) and then rinsed repeatedly with sterile distilled water and vortex agitation. For bacterial inoculations, PGP strains were grown in LB broth for 48 h at 28 °C. Cells were collected by centrifugation (12,000 g, 10 min), washed twice with 0.9% NaCl sterile solution, and adjusted to 10⁹ CFU mL⁻¹. The seeds were added to suspensions of the selected bacterial strain. After inoculation, 12 seeds of each were dispersed equidistantly on a paper towel previously moistened with 10 ml of sterile distilled water and were deposited in sterile petri dishes of 11.8 cm diameter. Plant seeds were covered by another paper towel dampened with 10 ml water and the petri dishes were incubated at 30 °C in a growth chamber in the dark. The numbers of germinated seeds of each were counted to determine their germination percentages. The germination time (GT, number of days post planting) was calculated using the following formula: GT = ∑(Gi × i)/∑Gi, where i is the number of days between seed sowing (day 0) and seed germination; and Gi is the number of seeds germinated on day i (Zhang et al., 2014).

Green house pot experiments

The effects of the three PGP bacterial strains on soybean, faba bean, wheat, pea, Chinese cabbage, and A. adenophora growth were also determined. The soil used for the potted trials was taken from the campus garden of Kunming University. The soil was air dried initially for 2 weeks and sieved to a particle size of <2 mm. The physicochemical properties of the soil were as follows: pH 7.06; total organic carbon 45.46 g kg⁻¹; total nitrogen 4.53 g kg⁻¹; available P 16.89 mg kg⁻¹; and available potassium 130.39 mg kg⁻¹. Polyvinyl chloride (PVC) pots that measured 15 cm in diameter and 40 cm in height were used. Each pot was filled with 8 kg of soil and watered to 2/9^th of their maximum holding capacity 24 h prior to sowing.

Three holes were made in the soil surface in each pot and one seed that had been inoculated with bacteria was gently planted in each hole. The holes were then filled. Pot experiments were conducted in a greenhouse (altitude 1,890 m; 24°58′N; 102°48′E) at Kunming University from March to August 2020. The temperature in the study area varied between 18.3 °C and 28.8 °C. On the 15^th day after seed germination, the least vigorous of the plants were removed from their respective planters. A root dip inoculation with 10 ml of each suspension of the PGP bacterial strains containing approximately 10⁹ cells per mL was performed. Pots were watered at 1/9^th of their maximum water retention capacity every 48 h after seed germination until the end of the experiments. Six replicate experiments (pots) were used for each treatment. The same number of replicates with uninoculated seeds were performed as controls. Root dips were performed with 10 mL sterilized water on the 15^th day after seed germination. Data, including aboveground and belowground heights and dry weights, were collected 60 days after seed planting. The data were analyzed for each plant. In order to determine the dry weight, the plant components were placed into a hot air oven at 60 °C until a dry weight was established.

Physicochemical analysis of soil samples

Physiochemical analyses of soil samples were carried out following procedures described by Kong et al. (2017). Briefly, soil organic matter composition was determined using the K₂Cr₂O₇-H₂SO₄ oxidation method (Nelson & Sommers, 1996). Total nitrogen levels were measured using the Kjeldahl method (ISO 11261, 1995). The available phosphorus and potassium levels were estimated using methods described by Kuo (1996) and Helmke & Sparks (1996), respectively. The pH (1:2.5 solution of soil to water) was measured using a pH meter (Mettler-Toledo International Inc., Shanghai, China).

Statistical analysis

Mean values and the least square difference of the relative nifH gene expression level, the IAA and ACC deaminase concentrations, the relative germination rate (GR) and GT, the relative aboveground and belowground heights, and the dry weights were calculated using SPSS (version 16.0; SPSS, Inc., Chicago, IL, USA). The results were expressed as mean ± standard error for each treatment. Relative GR, GT, aboveground and belowground heights, and dry weights were calculated as percentages of the values of individual germination and growth parameters measured for each individual plants in the treatment groups divided by the value measured for its plants in the control groups. Significant differences in these parameters between the treatment groups and their control groups were determined using the Kruskal–Wallis test with SPSS (version 16.0). Significant differences were established at P < 0.05. The correlation between the invasion years of A. adenophora to the genus and strain number of its rhizosphere PGP bacteria was analyzed using corrplot in RStudio (RStudio Team, 2017).

Isolation of bacteria from the rhizosphere soils of weed A. adenophora

To ensure a comprehensive screening of PGPR from the rhizosphere soil of the weed A. adenophora, 1,200 bacterial stains were isolated from the three rhizosphere soil samples (400 strains from each soil sample) belonging to A. adenophora located in three floristic areas (Fig. 1). The sample areas included Kunming with a south subtropical climate, Yuanjian with a middle subtropical climate, and Yunlong with a north subtropical climate in Yunnan, China (Table 1). The study site at Kunming had the longest A. adenophora invasion history (10–12 years) followed by that at Yunlong (6–8 years) and Yuanjiang (2–3 years). Each site had a different soil type (Table 1); the soil from A. adenophora in Kunming contained a higher (P < 0.05) level of organic C and lower level of NO₃⁻-N than those from Yuanjian and Yunlong. The soil from Yunlong had a higher (P < 0.05) pH value and contained a higher (P < 0.05) level of available P and a lower (P < 0.05) level of available K than those at Kunming and Yuanjiang (Table 2). Apart from pH, organic C, and available P and K, the rhizosphere soils of A. adenophora from the different locations did not differ significantly (P > 0.05) in any other parameters.

Screening for PGP bacteria from rhizosphere soils of A. adenophora from different floristic areas

Qualitative analyses using the selective media described above showed that 144 bacterial strains (Table S1) had at least one of the in vitro abilities for N-fixation, and IAA and ACC deaminase production (Table S1). To further identify those with the best PGP potentials we quantitatively analyzed each of their PGP abilities (Table S1). Soils from the different floristic areas differed in the numbers and composition of their PGP bacteria (Fig. 2). Of the 400 isolates from each study site, the rhizosphere soil from Kunming contained the most PGP bacteria (98 strains) followed by the soil samples from Yunlong (53 strains) and Yuanjiang (23 strains) (Fig. 2). Of the three PGP traits examined at each site, the IAA-producing bacteria were the most abundant (66 strains) in the soil at Kunming, while the N-fixing bacteria (28 strains) and the ACC deaminase-producing bacteria (11 strains) were the most abundant in the soils at Yunlong and Yuanjiang, respectively (Fig. 2).

Phylogenetic composition and distribution of the PGP bacteria at the rhizosphere soils of A. adenophora at different floristic areas

The 16S rRNA gene sequence data revealed that a majority (62.5%) of the PGP bacteria identified were bacteria that were closely related (≥97.6% 16S rRNA gene sequence similarity) to members of the genera Pseudomonas (27 strains), Providencia (20 strains), Chryseobacterium (14 strains), Ensifer (12 strains), Enterobacter (nine strains), and Hafnia (eight strains) (Fig. 3, Table S1). Their abundances at the different study sites also differed. A total of six, five, and 15 Pseudomonas strains were found at Kunming, Yunlong, and Yuanjiang, respectively, while three Enterobacteria were recovered from each of the sites. A total of 10 and four Chryseobacterium strains, and two and 10 Ensifer strains, respectively, were recovered from the study sites of Kunming and Yunlong. Providencia (20) and Hafnia (eight) strains were found only at the Kunming site (Table S1).

Correlation of invasion history of A. adenophora to the diversity and abundance of its rhizosphere PGP bacteria

The genus number and strain number of the PGP bacteria isolated from the rhizosphere soil of A. adenophora corresponded to the invasion history of A. adenophora. Thus, the study site of Kunming with a 10–12 years’ invasion history had the largest number (98 strains) and the most diverse (at least 19 genera) PGP bacteria, followed by that of Yunlong (65 strains, 15 genera) which had a 6–8 years’ invasion history. Yuanjiang had the fewest and least diverse bacteria (23 strains, 10 genera) with a short 2–3 years’ invasion history. Correlation analyses showed that significant positive correlations exist between the invasion years of A. adenophora and the genus number (P = 0.019) and the strain number (P = 0.022) of the rhizosphere PGP bacteria.

Effects of the PGP bacterial strains on seed germination of crops

PGP strains KM_A34 of Pantoea agglomerans with the highest relative nifH gene expression level (1.1028), KM_C04 of Enterobacter asburiae with the highest IAA production ability of (35.02 mg L⁻¹), and KM_A57 of Pseudomonas putida with the highest ACC deaminase production ability (9.23 U mg⁻¹) (Table S1) were chosen. Their influence on the germination and growth of soybean, faba bean, pea, wheat, and Chinese cabbage, as well as A. adenophora was evaluated. Seed inoculation with the three PGP bacterial strains had different effects on the GR and GT of the crop seeds examined. Treatment with the N-fixation strain, KM_A34, significantly (P < 0.05) increased the GR of the soybean, faba bean, and pea seeds (Fig. 4A), while the application of the high ACC deaminase-producing strain, KM_A57, significantly (P < 0.05) increased the GR of the pea, Chinese cabbage, and A. adenophora. Treatment with the IAA-producing strain, KM_C04, significantly (P < 0.05) increased the GR of the faba bean, pea, and Chinese cabbage, while the N-fixation strain, KM_A34, significantly (P < 0.05) reduced the GT of soybean seeds (Fig. 4B). The application of the IAA-producing strain, KM_C04, significantly (P < 0.05) reduced the GT of the faba bean, wheat, and A. adenophora seeds, and increased that of the soybean and pea (Fig. 4B). The application of the ACC deaminase-producing strain, KM_A57, significantly (P < 0.05) reduced the GT of the faba bean, wheat, and A. adenophora seeds, and increased the GT of the soybean and pea seeds (Fig. 4B).

Effects of the PGP bacterial strains on crop growth

The 60-day greenhouse pot experiments were used to evaluate the effects of seed bacterization and root dip inoculation with the three PGP bacterial strains on the growth of the same crops examined above (Fig. 5). In comparison with the controls, the ACC deaminase-producing strain, KM_A57, significantly (P < 0.05) increased the crop’s aboveground height (pea) (Fig. 5A) and dry weight (pea) (Fig. 5C), and the belowground height (wheat) (Fig. 5B) and dry weight (pea) (Fig. 5D). The N-fixation strain, KM_A34, significantly (P < 0.05) increased the crops’ aboveground height (pea and Chinese cabbage) (Fig. 5A) and dry weight (pea) (Fig. 5C), and their belowground height (Chinese cabbage) (Fig. 5B) and dry weight (pea) (Fig. 5C). The strain also decreased the belowground height (Fig. 5B) and dry weight (Fig. 5D) of the faba bean. The IAA-producing strain, KM_C04, significantly (P < 0.05) increased the crops’ aboveground height (soybean, pea, wheat, Chinese cabbage) (Fig. 5A), dry weight (soybean, pea, wheat, Chinese cabbage) (Fig. 5C), belowground height (faba bean, wheat, Chinese cabbage) (Fig. 5B), and dry weight (faba bean) (Fig. 5D). All three PGP bacterial strains had a significant (P < 0.05), positive effect on the aboveground height (Fig. 5A) and dry weight (except for a neutral effect with the IAA-producing strain) (Fig. 5C), the underground height (Fig. 5B) and the dry weight (Fig. 5D) of A. adenophora.