The native soil microbiome is critical for early root-associated microbiota assembly and canola growth

Abstract

Background

The native soil microbiome contributes to regulating the root-associated microbiota, root morphology, and plant growth. Using two canola (Brassica napus L.) genotypes contrasting in root size (small-rooted NAM23 and large-rooted NAM37), we investigated how the native soil microbiome influences canola establishment. Plants were grown in rhizoboxes containing gamma-irradiated (microbiome dysbiosis) or untreated (healthy microbiome) soils for 14 days. We evaluated plant growth and profiled bacterial and fungal communities in unplanted soil, rhizosphere soil, and root samples via DNA amplicon sequencing.

Results

Soil irradiation inhibited canola early growth, severely reducing shoot fresh mass (8 to 10-fold), root fresh mass and root length (3 to 13-fold). As expected, irradiation reduced microbial diversity and altered microbial community structure. The absence of significant soil physicochemical changes post-irradiation suggests that microbiome dysbiosis, rather than nutrient depletion, was the primary driver of plant growth suppression in irradiated soil. This growth suppression correlated with the depletion of potentially beneficial taxa (e.g., Sphingomonas, Alternaria prunicola, Fusarium, Gibberella avenacea, and Humicola nigrescens) and/or the enrichment of detrimental taxa (e.g., Mucilaginibacter, Leifsonia, and Trichoderma atrobrunneum) in both soil and roots. The large-rooted NAM37 outperformed the small-rooted NAM23 only in healthy microbiome-intact soils, but this growth advantage was not observed in unhealthy microbiome-disrupted irradiated soils.

Conclusions

Our findings directly demonstrate the critical role of a healthy soil microbiome in supporting canola establishment. The absence of growth disparities between genotypes in irradiated soil indicates that plant fitness is not attributed to fixed root phenotypes but a dynamic interplay between intrinsic root traits and the microbiome.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40793-025-00774-7.

Background

Soil ecosystems underpin plant growth by physical support, nutrient provision, and water retention, and their microbial communities act as keystone mediators of plant health and ecosystem function. Harboring up to 10 billion microorganisms per gram, soils host a dynamic interface where abiotic and biotic components interact to shape plant productivity [1]. The native soil microbiome serves as not only a reservoir of microbial functional diversity essential for nutrient cycling and soil health, but also as a primary inoculum for the root-associated microbiota, which plays pivotal roles in plant development and fitness [2, 3]. The assembly of these root-associated microbiota is governed by a complex interplay of abiotic soil properties (e.g., texture, pH, moisture, and nutrient availability) and intrinsic plant traits (e.g., genetic background, developmental stage, and physiological status), which collectively dictate microbial recruitment from soil to the rhizosphere and the root endosphere [4–6].

The soil and root-associated microbiota play a dual role in plant health, acting as both benefactors and potential antagonists. Beneficial microbes bolster plant immunity via systemic acquired resistance, induced systemic resistance, and antimicrobial metabolite production [7, 8], while coexisting pathogens can undermine this balance, impairing plant growth and productivity [9, 10]. This interplay among microbial populations dynamically mediates plant defense mechanisms, integrating immune responses with growth regulation [11]. The equilibrium of these interactions determines whether the microbiome promotes plant health or induces disease, which remains poorly resolved in natural systems.

To disentangle these interactions, reductionist approaches such as hydroponic and gnotobiotic systems have been employed to isolate microbial effects from soil complexity [12–14]. Axenic studies demonstrate that introducing specific plant growth promoting microorganisms can modulate the root-associated microbiota and benefit plant growth [12, 15]. Investigations under controlled conditions have also revealed how plant pathogen, root exudates, and nutrient availability can drive microbial colonization dynamics [16–20]. While these reductionist approaches provided valuable information on the complex interactions between the plant and its root-associated microbes, their artificial environments lack the ecological realism of native soil microbial networks, limiting the translatability to field conditions. Therefore, sterilized soil systems offer a more ecologically relevant context for root-associated microbiome studies [21, 22]. Traditional soil sterilization methods (e.g., autoclaving, UV, and chemical fumigation) induce substantial soil physicochemical alterations, whereas gamma irradiation uniquely achieves effective microbial depletion while preserving native soil physicochemical properties [23–25]. This key distinction of irradiation enables a rigorous assessment of the indigenous microbiome’s impact on root-associated microbial community assembly and plant growth.

A critical but overlooked dimension in plant-soil-microbe interaction is the role of inherent root phenotype. Root phenotypic traits, such as total root length, play a crucial role in shaping the rhizosphere microbiome [26]. A larger root system with enhanced root length and branching can create more heterogeneous microhabitats through increased exudate gradients and niche partitioning, thereby fostering microbial diversity and functional complexity [27, 28]. These root traits may amplify a plant’s capacity not only to assimilate resources [29, 30] but also to recruit beneficial microbes from intact soils, ultimately enhancing plant growth [31]. However, whether such growth advantages persist in disturbed soils where the diversity of microbial reservoirs is diminished, remains untested.

A robust and diverse soil microbiome forms the foundation of the root-associated microbiome, playing pivotal roles in plant growth and development [32]. To validate the importance of the native soil microbiome, most studies focus narrowly on pathogen suppression, with improved plant growth attributed to the elimination of soil-borne pathogens [33, 34]. Direct evidence linking soil microbiome integrity to root-associated community assembly and host phenotype—particularly across genotypes with divergent root traits—is lacking. To address this, we applied gamma irradiation to decouple native microbiome contributions in a fertile agricultural soil and evaluated responses in two canola genotypes with contrasting root sizes. Specifically, we hypothesized that: (1) soil disturbance by irradiation would reduce diversity and alter composition of the soil microbiome, cascading into dysbiosis in the rhizosphere and root-associated microbiomes; (2) impaired recruitment of a beneficial root-associated microbiota from irradiated soil would hinder shoot growth and root development; and (3) soil microbiome disruption would eliminate growth advantages of the large-rooted genotype NAM37, as a larger root system would fail to compensate for the absence of intact soil microbial reservoirs, rendering both genotypes comparably impaired. This study provides direct evidence that a functional native soil microbiome is essential for plant establishment, fundamentally recalibrating our understanding of plant-soil interactions.

Materials and methods

Soil preparation, genotype screening, and experimental setup

Field soil (silt loam, pH 5.5-6.0, 3.7% organic matter) was collected in 2018 from a research farm near Scott, SK (52°21’38.6"N, 108°50’00.8"W). Soil properties included: 2000 mg kg⁻¹ total nitrogen (N), 88 mg kg⁻¹ available N, 555 mg kg⁻¹ total phosphorus (P), 25 mg kg⁻¹ available Olsen P, 2225 mg kg⁻¹ total potassium, and 262 mg kg⁻¹ total sulfur (S). Soil was air-dried, sieved (< 2 mm), and gamma-irradiated at Nordion (Canada Inc., Québec) using a minimum dose of 50 kGy to eliminate most microorganisms while preserving soil properties. Irradiation effectiveness was evaluated by comparing microbial communities in gamma-irradiated and untreated soils through culture-based assays, soil respiration, and DNA amplicon sequencing (Method S1).

Two canola (Brassica napu L.) genotypes, NAM23 (small-rooted) and NAM37 (large-rooted), were selected from a nested-association mapping panel [35] (Method S2). The experiment was conducted in a growth chamber using rhizoboxes (44.3 × 28 × 0.3 cm, height × width × thickness) filled with 420 g of dry soil at a bulk density of 1.16 g cm⁻³. Surface-sterilized canola seeds (65% ethanol for 5 min, 1.2% sodium hypochlorite for 5 min, followed by 5 rinses with sterilized distilled water) were germinated on moist germination paper for 4 days. Seedlings with 3–4 cm roots were transplanted into the center of rhizoboxes (one seedling per rhizobox). The experimental design included 36 rhizoboxes (Method S3):

1. Planted: 2 soil irradiation conditions (irradiated vs. untreated) × 2 genotypes × 6 replicates = 24 rhizoboxes.

2. Unplanted: 2 soil irradiation conditions (irradiated vs. untreated) × 6 replicates = 12 rhizoboxes.

Rhizoboxes were positioned on a rack at 30° angle from horizontal, with the imaging panel oriented downward to optimize visibility of root growth. To maintain consistent moisture, rhizoboxes were irrigated daily with sterilized deionized water based on evaporation rates, determined by daily weight loss per rhizobox.

Sample processing and collection

On 14 days (T14) after transplanting, root images from both sides of each rhizobox were acquired using an Epson EU-88 scanner (Epson, USA) with total root length being quantified using WinRHIZO™ Tron software (Regents Instruments, Québec, Canada) (Method S2).

After imaging, 24 rhizosphere soil and 24 root samples were collected from planted rhizoboxes. More specifically, rhizoboxes were disassembled and plant shoots were separated from the root systems using a sterilized scalpel, with both fresh and dry shoot weights being recorded. Roots were shaken to remove loosely attached soil, transferred to a 250 ml Erlenmeyer flask containing 100 ml of 0.05 M NaCl solution and agitated on an orbital shaker at 180 rpm for 15 min at room temperature. Roots were rinsed five times with autoclaved distilled water before being cut into 0.5 cm segments. Rhizosphere soil was recovered by centrifugation at 3000 rpm for 15 min, with the supernatant discarded. Rhizosphere and root samples were sub-sampled into 2 ml Eppendorf tubes and stored at -80 °C for DNA extraction.

Unplanted soil (in total 28 samples, with 8 at T0 and 6 at T14 for both untreated and irradiated soils) were homogenized, sieved (< 2 mm), sub-sampled into 2 ml Eppendorf tubes and stored at -80 °C for DNA extraction and microbial profiling. In addition, unplanted (both at T0 and T14) and rhizosphere (at T14) soils were sieved (< 2 mm), air-dried for total and available N and P, total organic carbon, and total S analyses (Method S4).

DNA extraction and amplicon sequencing

Total DNA was extracted from 250 mg unplanted and rhizosphere soil using the DNeasy PowerSoil Pro Kit (QIAGEN, Germany), and root-associated microbial DNA from 100 mg fresh roots was extracted using the DNeasy Plant Pro Kit (QIAGEN, Germany).

Genomic DNA in the extracts was quantified using a Qubit 4 fluorometer (Invitrogen™, CA) and standardized to 5 ng ul^− 1 prior to PCR amplification. The bacterial 16 S rRNA gene was amplified using primers 342 F (5′-CTACGGGGGGCAGCAG-3′; 16 nt) and 806R (5′-GGACTACCGGGGTATCT-3′; 17 nt) [36], while fungal ITS regions were amplified using ITS1-F_KYO1 (TCG TCG GCA GCG TCA GAT GTG TAT AAG AGA CAG CTH GGT CAT TTA GAG GAA STA A) and ITS2-R_KYO2 (GTC TCG TGG GCT CGG AGA TGT GTA TAA GAG ACA GTT YRC TRC GTT CTT CAT C) primers [37]. Targeted bacterial and fungal regions were PCR-amplified using a T100™ Thermal Cycler (Bio-Rad, CA, USA) followed by purification. Purified products underwent library preparation and were sequenced on an Illumina MiSeq platform (Illumina Inc., CA, USA) [38].

Identification of colonies growing in the gamma-irradiated soil

Cells from colonies observed in gamma-irradiated rhizoboxes were transferred to 1/10 strength Tryptic Soy Agar (TSA) for bacteria and Potato Dextrose Agar (PDA) for fungi, followed by streaking to isolate single colonies. Bacterial isolates were cultured in half-strength Tryptic Soy Broth (TSB), while fungal isolates were cultured on PDA plates. Microbial DNA was extracted using the DNeasy^® UltraClean^® Microbial Kit (QIAGEN, Germany), following the manufacturer’s protocol.

PCR amplification was performed using a T100™ Thermal Cycler (BIO-RAD, California). Primers 27F (5’-AGAGTTTGATCMTGGCTCAG-3’) and 1492R (5’-TACGGYTACCTTGTTACGACTT-3’) were used for bacterial 16 S rRNA amplification [39], and primers ITS1 (5’-TCCGTAGGTGAACCTGCGG-3’) and ITS4 (5’-TCCTCCGCTTATTGATATGC-3’) were used for fungal ITS amplification [40]. Target regions were PCR-amplified, purified, and commercially sequenced by Macrogen Inc. (Seoul, South Korea) using Sanger sequencing (Method S5).

Raw sequence data were processed in Geneious Prime software (Geneious, Auckland, New Zealand), including primer trimming, low-quality reads removal, reverse complements generation, and forward/reverse reads alignment at 93% similarity to create consensus sequences. Taxonomic classification was performed via Basic Local Alignment Search Tool (BLAST) searches in the National Center for Biotechnology Information (NCBI) database.

Bioinformatics and statistical analysis

Quality-controlled bacterial and fungal amplicons generated from Illumina Miseq runs were processed in QIIME2 v2022.8 [41]. Sequences were demultiplexed and denoised with DADA2, which truncated to generate representative Amplicon Sequence Variants (ASVs) [42]. Taxonomic classification of the generated ASVs was performed using the SILVA database (Silva 138) for bacterial 16 S rRNA [43] and the UNITE database (version 9.0) for fungal ITS sequences [44] (Method S5). The total reads per sample ranged from 13,674 to 49,517 for bacteria and 11,400–105,798 for fungi, yielding 2167 unique bacterial ASVs and 2239 unique fungal ASVs. Sequence data were deposited in the NCBI Sequence read archive (SRA) under Bioproject PRJNA1062716.

Prior to alpha diversity (Shannon diversity) assessment, samples were rarefied to the mean sequence depth [45], with a minimum of 30,000 reads per sample for bacteria and 12,500 for fungi. The effects of soil irradiation, sampling time, and plant genotype on alpha diversity were assessed via analysis of variance (ANOVA) followed by a false-discovery-rate-corrected post hoc test.

Taxonomic composition was analyzed based on relative abundance, with the 20 most abundant bacterial and 11 most abundant fungal phyla (> 1% relative abundance) visualized using ggplot2 [46]. Beta diversity was assessed via principal component analysis (PCA) and Aitchison distance matrices computed on centered log-ratio (CLR) transformed data [47]. Microbial community structure was evaluated using permutational analysis of variance (PERMANOVA): effects of soil irradiation and sampling time were tested for unplanted soil, while effects of soil irradiation and plant genotype were analyzed for rhizosphere and root communities using the “adonis2” function [48].

Differential abundance in response to soil irradiation was determined using the Analysis of Compositions of Microbiomes with Bias Correction (ANCOM-BC2) for ASVs with relative abundance higher than 2% [49]. Spearman’s correlations between log-transformed ASVs and plant performance metrics (e.g.: shoot fresh and dry weight, root fresh weight, root/shoot weight ratio, and total root length) and soil properties (soil total and available N and P) were calculated using the “rcorr” function in the Hmisc package [50] and visualized in corrplot package [51].

Plant performance metrics (plant shoot fresh and dry weight, root fresh weight, total root length and root/shoot weight ratio), as well as soil chemical traits (soil total and available N and P, total organic C, and total S), were compared between gamma-irradiated vs. control untreated soils. Specifically, plant and soil trait data were first examined for normality, and variables exhibiting skewed distributions were log-transformed. Then statistical analysis of variance (ANOVA) was conducted by the “aov” function, followed by Tukey’s post-hoc pairwise test (“TukeyHSD” function). Results for soil traits were summarized in a table, and plant traits were visualized using the Graphics package [52].

Results

Soil irradiation effectively reduced viable microorganisms but did not completely sterilize soils

To verify the effectiveness of the irradiation, unplanted soil was collected and cultured on growth media at the beginning of the experiment (T0). Culturable bacterial and fungal colonies were observed in untreated soils but were absent in irradiated soils (Fig. 1A-B). However, once soils were moistened, untreated soils exhibited immediate high respiration rates, whereas irradiated soils showed negligible initial activity. Microbial regrowth commenced in irradiated soils by day 4, with cumulative respiration rates surpassing unplanted controls by twofold by day 5 (p < 0.001; Fig. 1C).

Fig. 1.

Microbial community suppression in gamma-irradiated soils: (A) Bacterial colonies (1/10 TSA media) and (B) fungal colonies (PDA media) in control unplanted (CU) vs. gamma-irradiated unplanted (GU) soils; (C) Respiration rates of CU vs. GU soils collected at the beginning of the experiment (T0); (D) Colonies from CU vs. GU after moist incubation in rhizoboxes for 14d (T14; n = 6)

Sanger sequencing of gamma-irradiated soil identified seven bacterial taxa (four Bacillus spp., and one Enterobacter spp., one Curtobacterium spp., and one Paenibacillus spp.) and eleven fungal taxa (nine Penicillium spp., and one Gloeotinia spp., and one Talaromyces spp.). In some cases, colonies exhibited co-occurrence of a bacterium and a fungus within a single morphologically indistinguishable colony. These co-occurring organisms included three colonies of Bacillus - Penicillium, one colony of Curtobacterium - Penicillium, and one colony of Paenibacillus - Penicillium.

Soil irradiation decreased microbial diversity and altered microbial community structure and composition, irrespective of plant genotype

In unplanted soils, gamma irradiation reduced Shannon diversity of bacteria by 60% and fungi by around 50%, compared to the untreated controls at T14 (p < 0.05; Fig. 2). Principal component analyses (PCA) and PERMANOVA revealed distinct microbial communities between irradiated and control untreated soils and between T0 and T14 (Table S1), with PC1 and PC2 accounting for 35.9% of the variation in the bacterial community and 23.9% in the fungal community (Fig. 3). A significant time x soil treatment interaction (p = 0.001 for bacteria and p = 0.004 for fungi; Table S1) in unplanted rhizoboxes pointed to larger differences in the community structure of both bacteria and fungi at T14 than at T0.

Fig. 2.

Alpha Shannon diversity of bacterial and fungal communities in control unplanted (CU) vs. gamma-irradiated unplanted (GU) soils at T0 and T14, and in the rhizosphere and roots of two canola genotypes (NAM23 and NAM37) at T14. Different letters indicate significant differences (p < 0.05) as determined by ANOVA followed by Tukey’s test (n = 6)

Fig. 3.

Beta-diversity of bacterial and fungal communities based on Principal Component Analysis (PCA) in control unplanted (CU) vs. gamma-irradiated unplanted (GU) soils at T0 and T14, as well as in rhizosphere soils and roots of two canola genotypes (NAM23 and NAM37) at T14. PERMANOVA p-values indicate the effects of time, soil irradiation condition, or genotype on bacterial and fungal community structures (n = 6)

In planted soils, soil gamma irradiation reduced bacterial and fungal diversities in both the rhizosphere and roots, although the decrease of root bacterial diversity was only significant in NAM37 but not NAM23 (p < 0.05; Fig. 2). In the rhizosphere soil, PC1 and PC2 accounted for 46.9% of the variation in the bacterial community and 36% in the fungal community (Fig. 3). In root samples, PC1 and PC2 explained 45.9% of bacterial community variation and 27.9% of fungal community variation. Overall, PCA highlighted the substantial impact of soil irradiation on bacterial and fungal community structures in both rhizosphere and roots (p = 0.001; Table S1).

The concentration of total DNA in both the unplanted and rhizosphere irradiated soils were significantly lower than in control untreated soils (p < 0.05; Fig. 4, Figs. S5-S7). Microbial community composition revealed similar bacterial and fungal phyla in irradiated vs. untreated soils across all sample types, however, their relative abundance shifted. In general, the relative abundance of Firmicutes, Proteobacteria and Ascomycota were higher in irradiated soils, while Chytridiomycota showed higher abundance under untreated conditions.

Fig. 4.

DNA concentrations and relative abundance of bacteria and fungi in control unplanted (CU) vs. gamma-irradiated unplanted (GU) soils at T0 and T14, as well as in rhizosphere soils and roots of two canola genotypes (NAM23 and NAM37) at T14. Different colors refer to different bacterial or fungal phyla. DNA concentrations in bold indicate significant differences (p < 0.05; n = 6)

In unplanted soils, the composition of dominant bacteria and fungi were affected by both irradiation and sampling time. Irradiation decreased relative abundance of Acidobacteriae (Acidobacteria), Chloroflexia (Chloroflexi), Alphaproteobacteria (Proteobacteria), Mortierellomycetes (Mortierellomycota), and Rhizophlyctidomycetes (Chytridiomycota), while increasing relative abundance of Bacilli (Firmicutes) compared to untreated controls at both T0 and T14 (Fig. 4, Fig. S5). At T14, Firmicutes and Ascomycota dominated the bacterial and fungal communities, with their relative abundances exceeding 80%, compared to lower levels observed at T0 (Fig. 4).

In planted soils, differences between bacterial and fungal compositions of both the rhizosphere soil and roots were greater between irradiated vs. untreated conditions than between NAM23 vs. NAM37. Irradiation reduced the relative abundance of Acidobacteriae (Acidobacteriota), Actinobacteria (Actinobacteriota), Gemmatimonadetes (Gemmatimonadota), Ktedonobacteria (Chloroflexi), Dothideomycetes (Ascomycota), Agaricomycetes (Basidiomycota), and Rhizophlyctidomycetes (Chytridiomycota), while increasing the relative abundance of Betaproteobacteria and Gammaproteobacteria (Proteobacteria) and Bacilli (Firmicutes) in both the rhizosphere and roots (Fig. 4, Figs. S6-S7). In rhizosphere soils, irradiation decreased relative abundance of Alphaproteobacteria, Chloroflexi, Thermoleophilia, Leotiomycetes, and Tremellomycetes, while in roots, it reduced abundance of Phycisphaerae, Saccharimonadia, and Sordariomycetes but increased Mortierellomycetes (Fig. 4, Fig. S6-S7).

Microbiome disruption from soil irradiation inhibited Canola growth and correlated with specific root-associated microbial taxa

Gamma irradiation drastically inhibited the growth of canola (NAM23 and NAM37), reducing shoot fresh weight by 8- to 10-fold, root fresh weight by 3- to 6-fold, and total root length by 10- to 13-fold compared to plants grown in control untreated soils (p < 0.05; Fig. 5). In unplanted soils, except for decreased soil available N in irradiated compared to untreated soils (T0 and T14; p < 0.05), other soil nutrients remained unchanged (Table 1). In planted soils for both genotypes, available N was lower while available P was higher in irradiated compared to untreated soils (Table 1).

Fig. 5.

Plant root systems (A), shoot fresh weight (B), root fresh weight (C), and root total length (D) for two canola genotypes (NAM23 and NAM37) in control untreated (C) vs. gamma-irradiated (G) soils grown in rhizoboxes for 14d (T14). Different letters indicated statistical significances (p < 0.05; n = 6)

Table 1.

Chemical analysis of soils from control unplanted (CU) and gamma-irradiated unplanted (GU) rhizoboxes at T0 and T14), and from planted rhizoboxes with Canola NAM23 (G23 and C23) and NAM37 (G37 and C37) at T14. Different letters indicate statistical significances (p < 0.05; n = 6)

Treatment	Rhizobox Type	Total N (mg kg− 1)	Available N (mg kg− 1)	Total P (mg kg− 1)	Available P (mg kg− 1)	Organic C (%)	Total S (mg kg− 1)
CUT0	Unplanted	2280a	102a	633a	34.5b	2.23a	410a
GUT0	Unplanted	2299a	56b	631a	34.8b	2.23a	411a
CUT14	Unplanted	2092a	133a	633a	39.9a	/	/
GUT14	Unplanted	2062ab	82c	619a	41.3a	/	/
C23T14	NAM23	2049ab	103b	621a	37.0b	/	/
G23T14	NAM23	2079ab	82c	633a	41.1a	/	/
C37T14	NAM37	2033b	99b	626a	36.2b	/	/
G37T14	NAM37	2073ab	85c	625a	40.3a	/	/

In addition to the general microbial community responses to irradiation (Figs. 2, 3 and 4), the abundance of specific microbial taxa changed, where differential abundance analysis of the irradiated vs. control soils revealed 57 bacterial and 57 fungal ASVs from unplanted soils, 52 bacterial and 37 fungal ASVs from rhizosphere soils, and 16 bacterial and 14 fungal ASVs from roots (p < 0.001; Fig. 6, Tables S2-S5). In rhizosphere soil and roots, microbes enriched in irradiated soil exhibited negative correlations with canola growth, while those predominant in untreated soils showed positive correlations (p < 0.01).

Fig. 6.

Log-fold change (LFC) abundance of differentially abundant bacterial (A) and fungal (B) ASVs in canola roots (p < 0.001): negative values (enriched in control soils) vs. positive values (enriched in gamma-irradiated soils). Spearman’s correlogram linking bacterial and fungal ASVs to soil/plant parameters: STN (soil total nitrogen), SAN (available nitrogen), STP (total phosphorus), SAP (available phosphorus), ShFw (shoot fresh weight), RFw (root fresh weight), RSh (root/shoot ratio), TRL (total root length). Red/blue circles indicated significant positive/negative correlations, with correlation coefficients displayed on the correlogram (p < 0.01; n = 6)

Specifically, root bacterial ASVs from Ktedonobacteraceae (ASV60, ASV63, ASV252, ASV68), Sphingomonas (ASV31), and Streptomyces (ASV243), as well as root fungal ASVs from Pleosporales (ASV27), Alternaria (ASV7), Fusarium (ASV13, ASV2), Gibberella (ASV11), Mortierella (ASV15), Metacordyceps (ASV36), Cladosporium (ASV83), and Humicola (ASV9) were more abundant in untreated compared to irradiated soils (p < 0.001). The abundance of these taxa correlated positively with shoot fresh weight and total root length (p < 0.01; Fig. 6, Tables S2-S3). In particular, similar trend was detected for bacteria from Ktedonobacteraceae and Pleosporales at genus level. In contrast, root bacterial ASVs from Mucilaginibacter (ASV1405), Leifsonia (ASV1416, ASV1407), Curtobacterium (ASV147), Methylobacterium (ASV4), and root fungal ASV5 (Trichoderma atrobrunneum) were enriched in irradiated soils, similar abundance trend was detected for bacteria Leifsonia and Methylobacterium at genus level (p < 0.001). Their abundance correlated negatively with shoot fresh weight (p < 0.01; Fig. 6, Tables S2-S3).

Rhizosphere communities mirrored this dichotomy: bacterial ASV31 and fungal ASVs (ASV7, ASV2, ASV13, ASV11, ASV15, ASV9) were enriched in untreated soils (p < 0.001), exhibiting positive correlations with shoot fresh weight, root fresh weight, and total root length (p < 0.01; Figs. S8-S9, Tables S4-S5). Conversely, bacterial ASVs (ASV1405, ASV1416, ASV1407, ASV147) and fungal ASV5 had higher abundance in irradiated soils (p < 0.001), correlating negatively with canola shoot fresh weight and total root length (p < 0.01; Figs. S8-S9, Tables S4-S5).

Disscussion

Impacts of soil disruption on soil and root-associated microbiota

The soil microbiome serves as the primary reservoir for plant root-associated microbiota. Thus, a robust and diverse soil microbiome plays foundational roles in shaping the plant microbiome [53]. Consistent with our first hypothesis, the rhizosphere and root-associated microbiomes of canola were reshaped by soil gamma-irradiation, decreasing microbial diversity and altering community composition (Figs. 1, 2, 3 and 4). Soil irradiation largely, but not completely eliminated culturable bacteria and fungi and reduced microbial respiration in unplanted soils (Fig. 1), indicating a remarkable decline in microbial community population and activity. Similar root-associated microbial diversity reductions were reported following soil autoclaving or gamma irradiation [54–57]. The observed reduction in microbial diversity is particularly significant, as higher soil microbial diversity is generally associated with enhanced microbial community function which contributes to key processes essential for plant growth, including nutrient cycling, disease suppression, and the regulation of plant hormones [58]. The decline in microbial diversity, therefore, likely disrupts these critical ecological functions, underscoring the broader impact of soil disturbance on plant-microbe interactions and soil health [59].

The observed increase in soil respiration after 4 days in the irradiation soils was likely due to a phenomenon referred to as the “Birch Effect”, where upon rewetting, cellular solutes or cell debris serve as carbon sources, thereby stimulating CO₂ release [60]. Further, while amplicon sequencing revealed structural shifts in microbial communities, it cannot exclude the interference of relic DNA from non-viable microbial debris in addition to live microorganisms [61]. This was apparent in our unplanted soils, where gamma irradiation reduced total microbial DNA concentrations and bacterial diversity at T0 and was more pronounced at T14, whereas fungal diversity remained stable at T0 but declined significantly by T14 (Fig. 3), suggesting the persistence of amplifiable extracellular DNA immediately post-irradiation. By T14, the collapse in microbial diversity and the decrease in DNA concentrations in gamma-irradiated soils (Fig. 6) indicated that extracellular DNA from dead cells was likely degraded after soil was moistened from the post-irradiation air-dried storage condition [62–64]. This was further supported by the reduced extractable DNA concentrations in irradiated soils between T0 and T14.

Notably, the reduction in microbial diversity was pronounced in the rhizosphere and roots, reflecting the dependence of root-associated microbiota on the soil microbiota as an inoculum reservoir [53]. The loss of microbial diversity was accompanied by shifts in community composition and abundance changes in specific taxa. In both rhizosphere and root samples, irradiation enriched stress-tolerant taxa such as Mucilaginibacter (ASV1405), Leifsonia (ASV1416 and ASV1407), and Trichoderma atrobrunneum (ASV5) (Fig. 6, Tables S2-S5). These taxa, which are often associated with disturbed environments, likely survived irradiation due to their resilience and subsequently colonized the vacant niches left by sensitive species [65, 66]. Conversely, irradiation depleted beneficial taxa such as Sphingomonas (ASV31) and Humicola nigrescens (ASV9) (Figs. 3 and 4), which are known to promote plant growth through nutrient solubilization and hormone regulation [67, 68].

The observed changes of soil and root-associated microbial diversity, composition, and structure post-irradiation align with previous studies showing that sterilization disrupts microbial community balance and functional redundancy, diminishing the capacity of soil microbiome to support plant health [58, 59, 69, 70]. Importantly, these effects were not limited to bulk soil (unplanted soil) but propagated into the rhizosphere and roots, highlighting the interconnection of soil and root-associated microbiota.

Impacts of root-associated microbial dysbiosis on plant growth

Plants grown in irradiated soils had reduced shoot fresh weight, root fresh weight, and total root length compared to those grown in untreated soils (Fig. 5). These growth defects were unlikely caused by nutrient limitation, as major soil nutrient levels remained sufficient for early canola growth (Table 1). The recommended rate is 119 lb ac^− 1 for achieving a yield of 50 bu ac^− 1, therefore, N levels at 112–164 lb ac^− 1 post-irradiation should be sufficient to satisfy early canola growth. This further verified that gamma irradiation had a notable advantage over conventional techniques such as autoclaving, fumigation, and dry heat, which often alter soil structure, organic matter, and nutrient availability through extreme thermal or chemical stress [23–25].

We thus assert that the observed growth inhibition stemmed from the disruption of the native soil microbiome, which impaired the assembly of a functional plant-associated microbial microbiota, adversely affecting plant health and productivity [71, 72]. These results validate our first two hypotheses and align with studies linking soil microbiome disruption to impaired plant development in Arabidopsis thaliana and in a cocultivation system of B. napus and Phelipanche ramose [69, 73]. Rhizosphere and root-associated microbial taxa enriched in untreated soils exhibited strong positive correlations with canola biomass and root development (e.g.: Sphingomonas, Alternaria prunicola, Fusarium, Gibberella avenacea, and Humicola nigrescens) (Fig. 6, Figs. S6-S9, Tables S2-S3), implicating their roles in nutrient mobilization, pathogen suppression, or phytohormone-mediated plant growth promotion. Among these taxa, Sphingomonas are known as plant growth-promoting microbes, which can produce auxins that have enhanced plant growth in tomatoes [67] and stimulated root development in B. napus [66]. H. nigrescens species have been identified as potent extracellular phytase producers, liberating soil inorganic phosphate in wheat [68], thereby contributing to improved phosphate availability and enhanced plant growth. Post-irradiation depletion of these taxa may lead to the loss of essential microbial symbionts that facilitate plant growth while concurrently perturbing the intricate metabolic networks within the soil microbiome that are critical for maintaining plant health.

Correspondingly, the prevalence of potential harmful microbes in irradiated soils may exacerbate plant growth defects. Microbes such as Mucilaginibacter, Leifsonia, Curtobacterium, Trichoderma atrobrunneum were enriched in both the rhizosphere and root samples, negatively correlating with canola shoot and root growth (Fig. 6, Figs. S6-S9, Tables S2-S5). Although some members from these taxa demonstrate potential in plant growth promotion [74–77], gamma irradiation can induce gene mutation of these microbes, shifting their ecological roles and activities in the plant-soil-microbe interactions [78]. Disruption of microbial-related nutrient cycling post-irradiation may also be involved as irradiation reduced available N in the rhizosphere by 14–20% (Table 1), potentially linked to the enrichment of denitrifying microorganisms, such as bacterial species from Massilia and Bacillus, and fungal species from Penicillium and Trichoderma in the rhizosphere soil (Figs. S8-S9, Tables S4-S5), which convert nitrate and nitrite to gaseous forms of N, including nitrous oxide and dinitrogen [79–81]. Soil disturbance via irradiation results in the elimination of most microorganisms, likely eradicating sensitive or beneficial microbial taxa while allowing more resilient, potentially detrimental taxa to survive and dominate. This disruption destabilizes the microbial equilibrium, impairs the recruitment of a healthy root-associated microbiota, and ultimately hinders plant growth and development [65, 66, 82].

Dominant role of native soil Microbiome over host regulation on root-associated microbiota

In addition to the primary microbiota resource from the native soil environment, the plant host also plays essential roles in regulating its root-associated microbiota via a wide range of strategies including genetic variation [83, 84], root exudate production [7, 85–87], and the regulation of plant physiological status such as the modulation of its immune system and nutrient uptake [88, 89]. Plant host is suggested to exert a stronger impact on the root endophyte compared to the rhizosphere, especially on the core microbiome [90]. Plant genetic material and the environment in which the plant grows (genotype x environment effects; G × E) interact to shape the root-associated microbiome [91–93]. In accordance with previous studies identifying dominant bacterial taxa from Actinobacteria and Proteobacteria and fungal taxa Mortierellomycetes, Sordariomycetes, and Tremellomycetes in the root-associated microbiome of canola [94, 95], our study similarly detected these major bacterial and fungal groups (Fig. 6, Figs. S6-S7).

Consistent with our third hypothesis, growth inhibition was observed in both canola genotypes in irradiated soil, while the large-rooted genotype (NAM37) exhibited a growth advantage only in untreated soils (Fig. 5). This suggests that absence of a healthy soil microbiome overrode the regulation of the large root phenotype on the assembly of the root-associated microbiota and plant growth. In untreated soils, the extensive root system of NAM37 with increased absorption area likely enhanced contact with beneficial microbes such as Sphingomonas and H. nigrescens, amplifying nutrient scavenging and growth promotion [68, 96]. However, soil disruption via irradiation eliminated beneficial microbe partners, rendering the potentially large-root phenotype irrelevant and underscoring the importance of the soil microbiome in supporting host-related plant growth advantages [97]. For example, corn genotypes lose rhizosphere microbiota distinctions when grown in autoclaved soil [98]. Our results extend this principle to root architecture, suggesting that breeding programs targeting root traits must consider plant-soil-microbe interactions to maximize agricultural benefits.

Conclusion

Our study directly demonstrates that a healthy native soil microbiome is essential for the assembly of a beneficial canola root-associated microbiota, which is in turn critical for plant growth and development. Disruption of the native soil microbiome via irradiation induced elimination of beneficial organisms, at the same time enriching root-associated microbes that were negatively correlated with plant growth. Although sterilization represents an extreme perturbation, serving as a model of dysbiosis, these findings highlight the critical role of fostering a diverse and robust native soil microbiome in sustaining plant health.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

BLH and LK designed the microbiome experiments and supervised the project. ML designed the experiment, conducted the research and performed data collection, analysis, and first draft of the manuscript. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

This study was supported by Canola Agronomic Research Program (CARP.2019.24), Natural Sciences and Engineering Research Council of Canada Discovery Grant (NSERC DG, RGPIN-2019-04158), Canada Excellence Research Chair (CERC), and China Scholarship Council (CSC).

Data availability

The datasets generated during this current study were submitted to the National Center for Biotechnology Information (NCBI) Sequence read archive (SRA) under Bioproject accession PRJNA1062716.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.