Diversity of rhizobia, symbiotic effectiveness, and potential of inoculation in Acacia mearnsii seedling production

Abstract

Black wattle (Acacia mearnsii) is a forest species of significant economic importance in southern Brazil; as a legume, it forms symbiotic associations with rhizobia, fixing atmospheric nitrogen. Nonetheless, little is known about native rhizobia in soils where the species is cultivated. Therefore, this study aimed to evaluate the diversity and symbiotic efficiency of rhizobia nodulating A. mearnsii in commercial planting areas and validate the efficiency of a potential strain in promoting seedling development. To this end, nodules were collected from four A. mearnsii commercial plantations located in Rio Grande do Sul State, southern Brazil. A total of 80 rhizobia isolates were obtained from black wattle nodules, and thirteen clusters were obtained by rep-PCR. Higher genetic diversity was found within the rhizobial populations from the Duas Figueiras (H′ = 2.224) and Seival (H′ = 2.112) plantations. Twelve isolates were evaluated belonging to the genus Bradyrhizobium, especially to the species Bradyrhizobium guangdongense. The principal component analysis indicated an association between rhizobia diversity and the content of clay, Ca, Mg, and K. Isolates and reference strains (SEMIA 6163 and 6164) induced nodulation and fixed N via symbiosis with black wattle plants after 60 days of germination. The isolates DF2.4, DF2.3, DF3.3, SEMIA 6164, SEMIA 6163, CA4.3, OV3.4, and OV1.4 showed shoot nitrogen accumulation values similar to the N + control treatment. In the second experiment (under nursery conditions), inoculation with the reference strain SEMIA 6164 generally improved the growth of A. mearnsii seedlings, reinforcing its efficiency even under production conditions.

Supplementary Information

The online version contains supplementary material available at 10.1007/s42770-022-00867-2.

Introduction

Acacia mearnsii De Wild., commonly known as the black wattle, is a forest species native to Australia belonging to the family Leguminosae (Fabaceae) and subfamily Caesalpinioideae [1, 2]. It is commercially used as raw material to produce tannins and wood chips in Africa and South America [3]. In southern Brazil, A. mearnsii forests occupy roughly 75 thousand hectares and are cultivated in monocultures and agroforestry systems, mainly to meet the production of tannins, chips, and pellets for exportation [4, 5]. In addition to being an economically important species, A. mearnsii also contributes to improving soil quality and sequestering carbon [6–8].

A feature that makes the genus Acacia highly adaptable to different environments is the ability to establish symbiotic association with rhizobia, which are responsible for the biological nitrogen fixation (BNF) process providing nitrogen for plant development, allowing the commercial cultivation of this tree to be performed in different regions of the planet. Acacia species are nodulated by rhizobia belonging to the genera Rhizobium, Mesorhizobium, Sinorhizobium, and Bradyrhizobium [9], in addition to Ochrobactrum, which is reported as able to nodulate Acacia mangium [10]. However, only a few studies have focused on the Acacia-nodulating rhizobial diversity. A. mearnsii is not a selective host [11], and it may be nodulated by Bradyrhizobium japonicum [9], Bradyrhizobium elkanii [12], Rhizobium tropici, and Rhizobium leguminosarum [13]. Nonetheless, Vargas et al. [14] observed that Bradyrhizobium sp. is the most common symbiont of A. mearnsii plants in southern Brazil, making all of their isolates belong to this genus, although their taxonomic status was not fully resolved. The authors also reported a wide variation in nitrogen fixation efficiency among the native isolates.

Inoculation with efficient rhizobia is an important practice in producing tree legume seedlings [15]. The use of this technology leads to greater seedling growth, and it is currently being developed for the African species of Acacia (i.e., A. mangium and Acacia auriculiformis) under nursery conditions [16–20]. In Brazil, the legislation allows the commercial production of inoculants containing two strains (SEMIA 6163 and 6164) for A. mearnsii [11], although the lack of tests in soil and fields led them to be removed from the list of official microorganisms in 2013 [21]. In other words, the former recommendation was solely based on testing under sterile conditions, and the possible effects of rhizobial inoculation on A. mearnsii commercial production remain unknown. As a result, inoculating seeds in commercial forest nurseries is not possible, albeit it can provide a series of benefits, including lower fertilization dependence, reduced costs, and the impact on the environment [14], higher seed germination speed [22], and improved forest quality [23], resulting in more remarkable survival and forest productivity [24, 25].

Given the above, this study aimed to (1) evaluate the diversity and establish the taxonomy status of A. mearnsii rhizobial isolates from southern Brazil, (2) determine the symbiotic efficiency of these native rhizobia, and (3) confirm, under nursery conditions, the efficiency of a putative elite strain (SEMIA 6164) on the quality of A. mearnsii seedlings.

Materials and methods

Rhizobial isolation and soil sampling

Root nodules of A. mearnsii trees and soil samples were collected from commercial plantations belonging to Tanagro S/A company in Rio Grande do Sul State (RS), southern Brazil, in Duas Figueiras (DF) in the city of Jaguarão (32°16′34 ″S 3°19′19″W), Seival (SE) in the city of Candiota (31°21′53″S 53°43′32″W), Camboatá (CA) in the city of Piratini (31°25′32″S 52°58′45″W), and Ouro Verde (OV) in the city of Cristal (31°06′30″S 52°02′15″W). At each sampling site and in a clonal plantation, four trees were selected to collect 40 nodules preserved in silica gel until processing in the Agricultural Microbiology Laboratory of the DDPA/SEAPDR in Porto Alegre (RS). Five nodules were randomly taken from each sampling point, surface disinfected in 70% alcohol for 1 min, 3% sodium hypochlorite for 3 min, and rinsed five times in sterile distilled water. The disinfested nodules were macerated with tweezers, seeded in Petri dishes with YMA medium, and incubated at 28 °C until well-isolated colonies could be observed [26]. The cultures were purified by successive subcultures on plates with a YMA medium and maintained in lyophilized cultures at SEMIA Rhizobia Collection (World Data Center on Microorganisms no. 443) of the SEAPDR. Surface sterilization of the nodules was confirmed by placing the water from the last wash in YMA. The reference strains SEMIA 6163 and 6164 were also acquired from the SEMIA Rhizobia Collection. Ten soil subsamples (0–20 cm layer) were collected and pooled together to obtain a representative soil sample from each site, which was analyzed for physicochemical properties according to Tedesco et al. [27] (Table 1).

Table 1.

Soil properties at the sampling sites

Sampling site	pH	Clay	OM	P	K	H + Al	Al	Ca	Mg
		------%------		---mg dm−3---		------------cmolc dm−3------------
Duas Figueiras (DF)	4.6	26	4.3	3.8	205.0	10.9	0.9	2.5	1.9
Seival (SE)	5.1	23	2.9	3.8	157.0	4.6	0.1	3.2	1.6
Camboatá (CA)	4.4	23	4.4	7.0	116.0	15.4	1.6	3.1	1.1
Ouro Verde (OV)	4.2	15	4.2	5.4	61.0	5.5	1.0	1.0	0.6

Total DNA extraction from rhizobia

The rhizobia were grown in a YM medium at 28 °C under constant agitation at 100 rpm for 48 h. Subsequently, DNA extraction and purification were performed using a standard protocol with phenol–chloroform [28]. To evaluate the quality of the extracted DNA, the samples were submitted to electrophoresis in 1.5% agarose gel.

Genotyping characterization

Rep-PCR reactions were performed using the enterobacterial repetitive intergenic consensus primers ERIC1-R (5′-ATG TAA GCT CCT GGG GAT TCA C-3′) and ERIC-2 (5′-AAG TAA GTG ACT GGG GTG AGC G-3′). Reactions were performed in a 25-mL volume containing genomic DNA (20–50 ng), 10 pmol of each primer, 10 mM of each dNTP, 1 mM of MgCl₂, 1 μL of DMSO, 1X Buffer (10 ×), and 1U of Taq DNA Polymerase (Thermo Scientific). A total of 39 cycles were performed: 94 °C for 5 min, followed by 37 cycles of 94 °C for 1 min, 50 °C for 1 min and 10 s, and 72 °C for 1 min; and a final extension of 72 °C for 5 min. The amplification products were subjected to electrophoresis in 1.5% agarose gel stained with Blue Green Loading Dye I (LGC Biotechnology). The band profile was analyzed using the GeljbdImages software. Genetic similarity among isolates was measured by Jaccard’s coefficient, and clusters were performed using the unweighted pair group method with arithmetic mean (UPGMA).

PCR amplification and partial sequencing of the 16S rRNA gene

The near-full length 16S rRNA gene was amplified using BacPaeF (5′AGAGTTTGATCCTGGCTCAG3′) and Bac1542R (5′AGAAAGGAGGTGATCCAGCC3′) primers in a final volume of 25 µL containing genomic DNA (20–50 ng), DreamTaq buffer 10 × (2.5 µL; Thermo Scientific), 10 mM dNTPs mix (0.5 µL), 10 µM of each primer (0.5 µL), DMSO (1 µL), and 0.1 µL of Dream Taq Polymerase (10 U/µL; Thermo Scientific). The PCR cycling program ran at 94 °C for 5 min, followed by 37 cycles at 94 °C for 1 min, 57 °C for 1 min and 10 s, and 72 °C for 1 min, followed by a final step at 72 °C for 5 min. Nucleotide sequences were determined on both strands of PCR amplification products using the Macrogen sequencing equipment (Macrogen Inc., Seoul, South Korea), an ABI3730XL, and sequencing primers 785F and 907R. Low-quality sequences were trimmed using the Chromas 2.6.4 software. Fragments were assembled into a single 1090–1462 bp sequence using the EMBOSS merger tool (http://www.bioinformatics.nl/cgi-bin/emboss/merger). Sequence identity was assessed by comparing 16S rRNA sequences of the SEMIA isolates with sequences from the EzBioCloud 16S rRNA server database (https://www.ezbiocloud.net/identify). The sequence data reported herein are publicly deposited at the GenBank under accession numbers MW131329–MW131339 and MW309504. The reference strains SEMIA 6163 and 6164 were sequenced by Menna et al. [12] and are deposited in GenBank under accession numbers AY904764 and AY904765, respectively.

According to the sequence accessions, a 16S rRNA phylogeny was reconstructed with the 62 strains of Bradyrhizobium species and Azorhizobium caulinodans NBRC 14845 ^T as an outgroup provided on LPSN (available at https://lpsn.dsmz.de/). The 16S rRNA sequences were aligned with the AlignSeqs function available via the ‘DECIPHER’ v2.14.0 R package [29]. The ‘phangorn’ v2.5.5 R package was used to select HKY + G + I (Generalized time-reversible with Gamma rate variation) as the best-fitting model and according to the Akaike information criterion using the modelTest function [30]. Then, ‘phangorn’ functions were also employed to construct phylogenetic trees using the maximum likelihood (ML) estimation. We first made a neighbor-joining tree and then fit the HKY + G + I ML tree using the neighbor-joining tree as a starting point. A protocol with 500 bootstraps was then computed; the tree was rooted in the outgroup. The R packages phytools and ggtree were used to view and edit the trees and cladograms [31, 32].

In vitro evaluation of indolic compound production

Indolic compound (IC) production—indole-3-acetic acid (IAA), indole-3-pyruvic acid (IPyA), and indole acetamide (IAM)—was evaluated according to Asghar et al. [33]. The isolates and reference strains were grown on King’s B medium supplemented with L-tryptophan (500 µg mL⁻¹) and incubated at 28 °C for 5 days. Afterward, the cultures were centrifuged, and 2 mL of the supernatant was added to 3 mL of Salkowski’s reagent (12 g L⁻¹ FeCl₃ + 7.9 M H₂ SO₄) and left for 30 min for color development. The reactions were read in a spectrophotometer (535 nm), and IC presence was estimated using a standard curve and expressed in micrograms per milliliter.

Diversity index

The Shannon–Weaver diversity index (H′) was estimated based on dendrograms of each sampled site (Figure S1) by counting the number of clusters at the 70% similarity level and the number of taxa within each cluster [34–36]. A principal component analysis (PCA) was used to verify the statistical correlation between soil properties and bacterial diversity [37]. Pair-wise squared Euclidean distances based on different soil properties were calculated for the four analyzed soils to obtain a double-centered distance matrix for factoring [38]. The H′ and PCA were computed using the PAST software [39].

Symbiotic efficiency evaluation of rhizobia

From the analyses of genetic diversity and the capacity to produce ICs, 12 isolates were selected, which, together with the reference strains, were tested for symbiotic efficiency in a greenhouse experiment. The A. mearnsii seeds used were provided by Tanagro S/A, and seed dormancy breaking was performed by immersion in water at 80 °C for 5 min [22]. For the planting of the seedlings, plastic tubs were used (55-cm³ volume) containing sterilized soil with the following physicochemical characteristics: pH = 5.4; exchangeable P = 42.9 mg dm⁻³; K = 41.0 mg dm⁻³; organic matter = 4.0 g kg⁻¹; clay = 7.0 g kg⁻¹; Ca = 1.1 cmol_c dm⁻³; and Mg = 0.5 cmol_c dm⁻³. The soil P and K levels were corrected, and then the sowing of A. mearnsii was performed. The treatments consisted of inoculating each isolate separately or the reference strain 7 days after sowing. The experiment was conducted with 4 repetitions and in a randomized block design. Inoculation was performed with 1.0 mL of the respective previously grown bacterial culture with a minimum concentration of 10⁹ CFU mL⁻¹. Two non-inoculated controls were included, one with added nitrogen fertilizer (N +) and one without (N −). At the beginning of the experiment, all treatments received a 9-mg dose of N in the form of ammonium nitrate. The N + treatment received two additional doses of N (36 mg each tube) 21 and 90 days after seedling emergence. The experiment was conducted for 60 days, at the end of which the shoot and root system of the plants were collected. Nodule dry weight (NDW), total plant height (TH), collar diameter height (CD), shoot dry weight (SDW), root dry weight (RDW), and shoot nitrogen accumulation (SNA) were evaluated. This last parameter was calculated by multiplying SDW by N concentration, which was determined with an organic elemental analyzer (Flash 200, Thermo Fisher Scientific).

An additional study was conducted with the reference strain SEMIA 6164 to prove its efficiency under field conditions and be recommended for commercial inoculant production. The study was conducted in a forest nursery located in Capela de Santana (29°39′59.5″S 51°20′46.1″W), RS, southern Brazil. The climate of this region is temperate-humid with an annual rainfall of around 1500 mm, well distributed throughout the year. The seeds were treated as previously described, and sowing was performed in 55-cm³ plastic tubs containing a mixture (1:1 v /v) of carbonized rice husk and soil. The soil used as substrate had the following characteristics: pH = 4.7; P = 2.0 mg dm⁻³; K = 44 mg dm⁻³; organic matter = 3 g kg⁻¹; clay = 430 g kg⁻¹; Ca = 0.4 cmol_c dm⁻³; and Mg = 0.9 cmol_c dm⁻³. An organomineral fertilizer 04–12-08 (2.22 g L⁻¹) was used as a base fertilizer.

The experiment had, as treatments, seedlings inoculated with SEMIA 6164 and seedlings without inoculation, grown in sterilized or non-sterilized soil in a 2 × 2 factorial arrangement, in a randomized block experimental design with four repetitions, each consisting of trays with 187 seedlings. Inoculation was performed as described previously, and the experiment was conducted for 100 days after sowing. At the end of this period, TH, CD, RDW, SDW, and NDW of the ten central seedlings of each tray were evaluated, and the TH, CD, RDW, and SDW data were used to calculate Dickson’s quality index (DQI), defined by Dickson et al. [40] as:

$$\mathrm{DQI}=\left(\mathrm{RDW},+,\mathrm{SDW}\right)/\left[\left(\mathrm{TH},/,\mathrm{CD}\right),+,\left(\mathrm{SDW},/,\mathrm{RDW}\right)\right]$$

where:

TH = seedling shoot height, from the collar to the tip of the tallest bud (cm);

CD = collar diameter height (mm);

RDW = root dry weight (g);

SDW = shoot dry weight (g).

The data from both experiments were submitted to analysis of variance (ANOVA) with the means compared by the Scott-Knott Test (p < 0.05) using the Sisvar software [41].

Results

Genotypic characterization of A. mearnsii-nodulating rhizobia isolates

A total of 80 rhizobia isolates were obtained from black wattle nodules and, together with the reference strains (SEMIA 6163 and 6164), were genotypically characterized by rep-PCR. Thirteen clusters were obtained (Fig. 1), seven of which were formed by up to three isolates. Three isolates showed a degree of similarity below 70% (DF2.3, SEMIA 6163, and DF3.3) compared to the others, occupying isolated positions. Nevertheless, 35 isolates (43%) showed a degree of similarity above 70% among themselves, forming in group L. The reference strains SEMIA 6164 and SEMIA 6163 clustered in groups C and D, respectively. No isolate showed the same Rep-PCR profile as the reference strains. The maximum similarity between one of the reference strains, SEMIA 6164, and one isolate was 76% (DF1.7 and CA2.2). Considering the number of individuals in each of the 13 clusters, the set of 82 rhizobia showed a Shannon index equal to 1.987, an evenness of 0.775, and an equitability of 0.775. An isolate was selected from each cluster, opting, whenever possible, for the one with the highest IC production to be evaluated for its phylogeny and symbiotic efficiency (Table 2).

Fig. 1.

Dendrogram of genotypic similarity of Acacia mearnsii-nodulating rhizobia from rep-PCR profiling estimated by UPGMA and Jaccard’s coefficient

Table 2.

Indole compound (IC) production by rhizobial isolates obtained from A. mearnsii nodules

Rhizobial strain	IC (µg mL−1)
SE4.4	46.62 ± 2.72
CA1.4	44.02 ± 2.34
CA3.7	38.43 ± 1.54
OV1.5	28.30 ± 0.07
OV1.4	27.92 ± 0.02
OV3.4	25.84 ± 0.70
CA4.3	24.28 ± 0.51
SE2.1	8.69 ± 2.76
DF2.3	0.00 ± 0.00
DF3.3	0.00 ± 0.00
DF2.4	0.00 ± 0.00
SE4.9	0.00 ± 0.00
SEMIA 6163	0.00 ± 0.00
SEMIA 6164	0.00 ± 0.00

Rhizobia phylogenetic analysis

The 16S rRNA sequences from the 12 selected isolates were used in similarity-based searches in a 16S rRNA curated sequence database (Table 3). Considering the arbitrary genus threshold of 94.5% for 16S rRNA gene identity [42], our isolates belong to the Bradyrhizobium genus. Bradyrhizobium rifense CTAW71^T was the top-hit taxon for DF3.3 (98.2% of similarity), whereas Bradyrhizobium guangdongense CCBAU 51649 ^T was the top-hit taxon for the remaining isolates (99.6–100% of similarity). The reference strain SEMIA 6163 had the highest similarity to Bradyrhizobium ganzhouense RITF806^T, while SEMIA 6164 had the highest similarity to Bradyrhizobium guangzhouense CCBAU 51,670.^T

Table 3.

16S rRNA sequence similarity analysis of Acacia mearnsii-nodulating rhizobial isolate strains

Name	Top-hit taxon	Top-hit strain	Similarity (%)	Completeness (%)
DF3.3	Bradyrhizobium rifense	CTAW71T	98.2	77.3
OV1.5	Bradyrhizobium guangdongense	CCBAU 51649 T	99.9	94.0
CA4.3			99.6	100.0
CA3.7			99.8	100.0
CA1.4			99.8	100.0
DF2.3			100.0	100.0
DF2.4			99.6	99.6
OV1.4			99.9	100.0
OV3.4			99.9	100.0
SE2.1			99.6	99.2
SE4.9			99.6	100.0
SE4.4			99.4	95.3
SEMIA 61631	Bradyrhizobium ganzhouense	RITF806T	99.85	96.6
SEMIA 61641	Bradyrhizobium guangzhouense	CCBAU 51670 T	99.86	100.0

¹Menna et al.[12]

A phylogenetic tree was also generated with the isolate sequences, reference strains (SEMIA 6163 and SEMIA 6164), and Bradyrhizobium-type strains (Fig. 2). The isolate DF3.3 was found grouped with low support along with Bradyrhizobium ganzhouense RITF806^T, Bradyrhizobium cytisi CTAW11^T, Bradyrhizobium rifense CTAW71^T, and SEMIA 6163. The remaining isolates were grouped along with Bradyrhizobium guangdongense CCBAU 51649 ^T, Bradyrhizobium manausense BR 3351 ^T, and SEMIA 6164, albeit with low support.

Fig. 2.

Phylogenetic tree of 16S rRNA gene sequences as inferred by ML. The black circles at the branching points indicate bootstrap values > 70%. Genbank IDs are shown within parentheses. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site

Diversity index

Higher genetic diversity was found within the rhizobial populations from the Duas Figueiras (DF) (H′ = 2.224) and Seival (SE) (H′ = 2.112) plantations. Camboatá (CA) (H′ = 1.982) had intermediary diversity, and Ouro Verde (OV) (H′ = 1.527) had the lowest diversity index. The PCA investigated the relationships between abiotic soil parameters and rhizobial diversity index H′ (Fig. 3). The first two PCA dimensions explained 89.48% of the total variance, with PC1 accounting for 64.86% and PC2 for 24.62% of the variance. The PC1 separated OM, Al, and P from the other variables, while PC2 grouped H′ with clay and the basic cations Ca, Mg, and K.

Fig. 3.

Principal component analysis (PCA) of the soil chemical characteristics and rhizobial Shannon–Weaver diversity index (H′) in the four sampling sites. DF Duas Figueiras, SE Seival, CA Camboatá, OV Ouro Verde

Symbiotic efficiency evaluation of rhizobia

The reference strains (SEMIA 6163 and 6164) and twelve isolates (SE2.1, SE4.4, SE4.9, OV1.4, OV1.5, OV3.4, CA1.4, CA3.7, CA4.3, DF2.3, DF2.4, and DF3.3) selected from the rep-PCR analysis and IC production capacity were tested for their symbiotic efficiency in the greenhouse. Both isolates and reference strains induced nodulation and fixed N via symbiosis with Acacia plants after 60 days of germination. Although inoculation with rhizobia did not influence CD, it significantly affected seedling TH (Table 4), ranging from 6.43 to 9.63 cm. Except for CA1.4 and CA3.7, all rhizobia promoted seedling height growth similarly to the control treatment with nitrogen fertilization. These, in turn, showed twice as much growth as seedlings from the treatment without inoculation and nitrogen fertilization.

Table 4.

Collar diameter (CD), total height (TH), nodule dry weight (NDW), shoot dry weight (SDW), root dry weight (RDW), and shoot nitrogen accumulation (SNA) in A. mearnsii seedlings inoculated with rhizobial strains

Treatment	CD	TH	NDW	SDW	RDW	SNA
	mm	cm	------------------------------mg plant−1------------------------------
SE2.1	4.3 ± 0.03 a	9.13 ± 0.58 a	5.63 ± 0.78 a	31.10 ± 4.59 b	14.60 ± 1.41 c	0.98 ± 0.13 a
SE4.4	4.6 ± 0.08 a	8.13 ± 0.77 a	7.93 ± 2.07 a	24.26 ± 6.39 b	19.13 ± 1.48 b	0.49 ± 0.08 b
SE4.9	4.6 ± 0.03 a	8.80 ± 0.49 a	4.20 ± 0.92 b	23.16 ± 0.17 b	11.53 ± 0.56 c	0.88 ± 0.05 b
OV1.4	4.0 ± 0.05 a	7.30 ± 0.85 a	4.46 ± 0.78 a	22.63 ± 4.58 b	11.13 ± 1.08 c	1.23 ± 0.18 a
OV1.5	4.6 ± 0.03 a	8.16 ± 0.14 a	5.60 ± 1.51 a	29.86 ± 1.41 b	13.56 ± 2.93 c	1.11 ± 0.07 a
OV3.4	4.0 ± 0.05 a	9.20 ± 1.42 a	4.60 ± 0.62 a	31.80 ± 4.90 b	13.56 ± 0.37 c	1.15 ± 0.17 a
CA1.4	3.3 ± 0.03 a	6.43 ± 0.24 b	3.10 ± 0.20 b	17.23 ± 1.23 b	8.50 ± 1.38 c	0.71 ± 0.14 b
CA3.7	4.0 ± 0.05 a	6.70 ± 0.61 b	5.43 ± 0.87 a	22.13 ± 2.76 b	15.26 ± 0.72 c	0.16 ± 0.03 b
CA4.3	4.3 ± 0.03 a	9.63 ± 0.98 a	5.63 ± 0.26 a	33.23 ± 2.83 b	15.70 ± 0.66 c	1.13 ± 0.23 a
DF2.3	4.0 ± 0.05 a	8.00 ± 0.63 a	3.70 ± 0.81 b	23.76 ± 1.16 b	14.00 ± 0.87 c	0.90 ± 0.18 b
DF2.4	4.0 ± 0.05 a	8.06 ± 0.41 a	6.63 ± 1.32 a	31.33 ± 2.23 b	12.43 ± 1.46 c	1.17 ± 0.08 a
DF3.3	5.3 ± 0.03 a	7.86 ± 1.20 a	3.56 ± 0.33 b	30.63 ± 3.56 b	10.60 ± 2.40 c	0.86 ± 0.14 b
SEMIA 6163	4.6 ± 0.03 a	8.66 ± 0.63 a	6.13 ± 0.69 a	33.70 ± 3.42 b	13.46 ± 1.70 c	1.30 ± 0.15 a
SEMIA 6164	4.6 ± 0.03 a	8.46 ± 0.52 a	5.36 ± 0.08 a	36.33 ± 2.08 b	13.83 ± 0.80 c	1.58 ± 0.01 a
N +	5.0 ± 0.00 a	9.70 ± 1.27 a	0.00 ± 0.00 c	48.00 ± 0.55 a	27.70 ± 7.30 a	1.26 ± 0.06 a
N-	2.6 ± 0.03 a	4.03 ± 0.92 c	0.00 ± 0.00 c	5.56 ± 0.44 c	10.56 ± 2.72 c	0.29 ± 0.02 c
F value	1.918	3.037	3.276	8.063	3.250	6.652
p value	0.0545	0.0032	0.0018	0.000	0.0019	0.000
Average	4.2	7.98	4.71	27.44	14.01	0.95
CV (%)	18.45	17.11	34.75	20.14	29.27	21.92

Means in each column followed by different letters differed significantly at p < 0.05 (Scott-Knott)

Rhizobia isolate inoculation also significantly differed in nodule dry weight (NDW). Most of the selected rhizobia isolates (SE2.1, SE4.4, OV1.4, OV1.5, OV3.4, CA3.7, CA4.3, and DF2.4) showed similar average NDW as both reference strains. Additionally, the isolate SE4.4 presented the highest NDW: 7.93 mg plant⁻¹ on average. In the control treatments, root nodule formation was not observed with and without N addition.

The SDW of the seedlings inoculated with the isolates tested did not differ from the reference strains. Although the SDW obtained with rhizobia inoculation was lower than the control treatment with nitrogen fertilization (48.00 mg plant⁻¹), it exceeded five times (27.93 mg plant⁻¹, in the averages of the 14 isolates) the treatment without inoculation and without nitrogen fertilization (5.56 mg plant⁻¹).

Similar to SDW, the highest RDW was obtained in the control treatment with nitrogen fertilization (27.70 mg plant⁻¹). Of the isolates tested, isolate SE4.4 showed the highest root dry mass (19.13 mg plant⁻¹) and was almost twice as high as the treatment without inoculation (10.56 mg plant⁻¹).

Rhizobia inoculation and nitrogen application significantly influenced N in seedling SNA; N content ranged from 0.42 to 1.39 mg plant⁻¹. The inoculation of isolate DF2.4 resulted in higher N content in the shoot of the seedlings (1.39 mg plant⁻¹), followed by SEMIA 6164 (1.37 mg plant⁻¹), whereas the control treatment with nitrogen application showed 1.20 mg plant⁻¹ of N. Nevertheless, the lowest content was obtained in the N − control treatment (0.42 mg plant⁻¹). The isolates DF2.4, DF2.3, DF3.3, SEMIA 6164, SEMIA 6163, CA4.3, OV3.4, and OV1.4 showed similar SNA values to the N + control treatment.

In the second experiment (under nursery conditions), we observed that, in general, inoculation with the reference strain SEMIA 6164 improved A. mearnsii seedling growth (Table 5). Inoculation with SEMIA 6164 promoted higher CD values by roughly 20.9%, TH by 24.5%, SDW by 13.03%, RDW by 10.31%, DQI by 8.3%, NDW by 16.8%, and SNA by 26.0% compared to the control treatment with no inoculation. It is important to note that the non-inoculated treatment showed nodule formation due to the presence of native rhizobia in the environment of the commercial seedling production nursery in which the experiment was conducted.

Table 5.

Analysis of variance for collar diameter (CD), total height (TH), shoot dry weight (SDW), root dry weight (RDW), Dickson quality index (DQI), nodule dry weight (NDW), and shoot nitrogen accumulation (SNA) in Acacia mearnsii seedlings inoculated with SEMIA 6164

Main effects	CD mm		TH cm		SDW mg plant−1			RDW mg plant−1		DQI			NDW mg plant−1		SNA mg plant−1
Non-sterilized soil	0.81 ± 0.03 a		7.24 ± 0.36 a		187.87 ± 5.04 a			87.25 ± 1.80 b		0.024 ± 0.00 b			8.41 ± 0.13 a		5.11 ± 0.22 a
Sterilized soil	0.82 ± 0.06 a		7.40 ± 0.60 a		176.27 ± 7.00 a			96.65 ± 2.91a		0.026 ± 0.00 a			7.39 ± 0.46 b		5.90 ± 0.54 a
Rhizobia inoculation	0.90 ± 0.03 a		8.11 ± 0.37 a		193.21 ± 4.78 a			96.46 ± 1.91 a		0.026 ± 0.00 a			8.45 ± 0.40 a		6.14 ± 0.50 a
No inoculation	0.74 ± 0.04 b		6.51 ± 0.43 b		170.93 ± 5.07 b			87.44 ± 2.92 b		0.024 ± 0.00 b			7.23 ± 0.12 b		4.87 ± 0.20 b
	F value	p value	F value	p value		F value	p value		F value	p value	F value	p value	F value	p value	F value	p value
Soil	0.027	0.8734	0.107	0.7506		3.884	0.0802		16.933	0.0026	5.153	0.0494	10.283	0.0107	3.150	0.1038
Rhizobia inoculation	11.482	0.0080	11.759	0.0075		14.305	0.0043		15.583	0.0034	5.153	0.0494	19.205	0.0018	8.350	0.0147
Soil × rhizobia inoculation	4.313	0.0676	4.375	0.0660		3.403	0.0982		3.062	0.1141	1.855	0.2063	12.145	0.0069	6.750	0.0248

Discussion

Sequencing data indicate that most isolates showed higher similarity to Bradyrhizobium guangdongense CCBAU 51649 ^T, a strain isolated from peanut (Arachis hypogaea) nodules [43, 44]. Isolate DF3.3 showed high similarity with B. rifense CTAW71^T initially described as nodulating Cytisus villosus, an African shrub species [45]. SEMIA strain 6163 was highly similar to B. ganzhouense RITF806^T, considered an efficient fixer in symbiosis with Acacia melanoxylon [46]. SEMIA 6164 was highly similar to B. guangzhouense CCBAU 51670 ^T, a peanut nodulating strain [47]. A high diversity of Bradyrhizobium species nodulating A. mearnsii is not surprising because, as reported elsewhere, soils in southern Brazil show a wide diversity of nifD haplotypes associated with Acacia spp., as a result of the introduction of these forest species into the region [48].

However, as evidenced by our phylogenetic tree, the 16S rRNA gene analysis is not enough to define bacterial species, especially those from the Rhizobiales family [49]. Trees based on a few thousand nucleotides, such as those based on a single phylogenetic marker (e.g., the 16S rRNA gene), tend to have branches with low bootstrap values [50]. In the phylogenetic analysis, the isolates and reference strains were found to be grouped with high support with in a clade that contained Bradyrhizobium americanum, B. amphicarpaeae, B. arachidis, B. betae, B. cajani, B. canariense, B. centrosematis, B. cytisi, B. daqingense, B. denitrificans, B. diazoefficiens, B. frederickii, B. ganzhouense, B. guangdongense, B. guangxiense, B. huanghuaihaiense, B. ingae, B. iriomotense, B. japonicum, B. kavangense, B. liaoningense, B. lupini, B. manausense, B. nanningense, B. niftali, B. nitroreducens, B. oligotrophicum, B. ottawaense, B. rifense, B. shewense, B. stylosanthis, B. subterraneum, B. symbiodeficiens, B. vignae, and B. yuanmingense, evidencing the low resolution of the 16S sequence analysis. Thus, the isolates and strains may belong to any of these species or represent new species within the same clade.

In addition to genotypic characterization by rep-PCR, we used IC production as a selection criterion for rhizobia that would be evaluated for symbiotic efficiency with A. mearnsii. This criterion was adopted because, in legumes, nodule formation and biological nitrogen fixation can be enhanced by the presence of IC [51]. The synthesis of IAA by rhizobia can be induced by flavonoids, interfering with nodule number and mass and biological nitrogen fixation, and it is critical for establishing symbiosis by some strains [52]. However, our results showed that this does not seem to be a general rule. Most of the analyzed parameters (CD, TH, NDW, SDW, and RDW) were not affected by IC production by rhizobia. Moreover, IC production negatively correlated with SDW (r =  − 0.534; p = 0.049) and SNA (r =  − 0.578; p = 0.0305). For A. mearnsii, IC production by rhizobia does not appear to be such an advantageous trait.

Duas Figueiras and Seival were the sites that showed higher rhizobia diversity compared to Camboatá and Ouro Verde. These differences may be linked to the chemical characteristics of the soil, more specifically pH and clay, Ca, Mg, and K levels. Rhizobia diversity is affected by several factors, including soil fertility, climatic characteristics, and management practices [53–55]. In this context, soil pH is considered the factor that most affects the diversity of rhizobia in soil [56–58], since many rhizobia are unable to survive under acidic conditions [59] thereby impacting their diversity [60–62]. Our findings are similar to those reported by Liu et al. [63], who observed that soil pH and Mg²⁺ and Ca²⁺ levels strongly influenced the β-diversity of the rhizobial community in the rhizosphere of soybean.

In addition to soil chemical properties, different land-use backgrounds, with crop rotation or monoculture of the host crop, can also affect the diversity of rhizobia populations. The Duas Figueiras and Seival areas were cultivated for the first time with A. mearnsii, while the Camboatá and Ouro Verde areas have been cultivated at least four consecutive times with the species. Thus, planting with A. mearnsii at the same site for several production cycles may have intensified the selection of specific groups of rhizobia, consequently reducing diversity. Authors such as Coutinho et al. [64], Zilli et al. [65], and Nkot et al. [66] reported a reduction in the diversity of rhizobia populations in the soil due to successive cultivation with the same crop, resulting from the selection of specific rhizobia taxa. Furthermore, Shao et al. [67] observed that growing peanuts in monoculture for long periods reduced the symbiotic rhizobial biodiversity; this effect may influence plant growth. In our study, the highest rhizobia diversity occurred in areas with the highest diameter growth of A. mearnsii trees. In areas with higher diversity (i.e., Duas Figueiras and Seival), the diameter at breast height (DBH) was 17.7 and 16.5 cm, respectively, while the DBH of Camboatá was 14.5 cm and Ouro Verde was 11.8 cm, showing a positive correlation between DBH and H′ (r = 0.977; p = 0.0227). These results are similar to those obtained by Yan et al. [68], who indicated rhizobial diversity as an indicator of productivity due to the high correlation between productivity and rhizobial diversity. Hence, the rhizobial diversity index can be considered a good indicator to predict the growth of A. mearnsii plantations. At the same time, our findings reinforce the importance of inoculation, even in areas traditionally cultivated with the species, because the most dominant rhizobia do not seem to be the most efficient.

The greenhouse experiment showed that inoculation of the isolates CA4.3, OV1.4, OV3.4, DF2.3, DF2.4, DF3.3, and the strains SEMIA 6163 and SEMIA 6164 increased A. mearnsii seedling development compared to the N − treatment. Of note, inoculation with DF2.4 and SEMIA 6164 increased the SNA values by roughly three times compared to N − treatment, demonstrating that these rhizobia are highly efficient in symbiosis with A. mearnsii. Our results corroborate Vargas et al. [14], who observed that SEMIA 6164, together with two prospected isolates, was an efficient N − fixer for A. mearnsii seedlings. Thus, the isolate DF2.4 and the SEMIA 6164 strain proved to be highly promising alternatives for nurseries producing A. mearnsii seedlings. Some isolates did not show the same efficiency in increasing SNA values, although they showed growth (TH and SDW) similar to the most efficient isolates and strains. The lower symbiotic efficiency may have been compensated by IC production in some of these isolates; the ICs assist in cell division, cell extension and differentiation, root formation, and vegetative growth control processes [69].

The results obtained under nursery conditions confirm the effectiveness of SEMIA 6164, as indicated in our greenhouse experiment and previously reported by Vargas et al. [14]. Overall, inoculation with SEMIA 6164 increased growth (TH and CD), dry mass (SDW and RDW), and SNA compared to the treatment without inoculation (Table 4). Similar findings were reported for other forest species under nursery conditions. For instance, Sutherland et al. [70] observed that rhizobia inoculation promoted substantial gains in nine African species of the genus Acacia. Brockwell et al. [24] observed that inoculation of A. mearnsii seeds with Bradyrhizobium sp. led to more vigorous and well-nodulated seedlings. In addition, Salto et al. [15] noted that the use of inoculum containing three rhizobium genera (Mesorhizobium sp., Bradyrhizobium sp., and Sinorhizobium sp.) in Prosopis alba seedlings resulted in higher nodule numbers and diameter at collar height (DC). Karthikeyan and Arunprasad [23] found that Rhizobium aegyptiacum inoculation increased by more than seven times the shoot dry mass, four times the root dry mass, and five times the nitrogen content in Pterocarpus santalinus seedlings.

Inoculation with rhizobia benefits A. mearnsii since seed germination and remains throughout tree growth in the field. São José et al. [22] observed that seeds inoculated with SEMIA 6164 showed higher germination speed than non-inoculated seeds. Brockwell et al. [24] observed that A. mearnsii seedlings inoculated with rhizobia had a 12% increase in survival rates and a 21% increase in individual volume at three years of age.

Conclusions

In this context, our findings reinforce the importance of inoculation as a practice that must be adopted in commercial nurseries of A. mearnsii seedlings. Above all, they proved the efficiency of SEMIA 6164, gathering information that will be useful for the Ministry of Agriculture, Livestock, and Food Supply of Brazil to include it in the list of strains authorized for the production of inoculants on a commercial scale, finally allowing seedling producers access to the technology [21].

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

Material and methods preparation, and analysis were performed by JFBSJ, MASH, CGV, CB, and BBL. AAS and JO performed field collections and logistical support. Grant and fellowships acquisition were provided by JFBSJ, LMPP, and AB. The manuscript was written by LKV, JFBSJ, and AB. All authors commented on previous versions of the manuscript and approved the final manuscript.

Funding

This work was financed by a grant and fellowships from the Fundação de Amparo à Pesquisa do Estado Do Rio Grande do Sul (Fapergs/Brazil) and Fundo Estadual de Desenvolvimento Florestal- FUNDEFLOR (SEAPDR/Brazil).

Declarations

Conflict of interest

The authors declare no competing interests.