Microbial dynamics in rice ecosystem under supplementation of organic sources of nitrogen with inorganics and their concomitant impact on yield

Abstract

Microorganisms play crucial role in the ecosystem building. Their presence or absence in a particular environment are indicative of the web of interactions they undergo to impact the life of other components of the system. The current experiment was undertaken in rice-rice cropping sequence, for two years, to understand the changes in composition of microorganism as a result of interaction of herbicide and different nutrient sources added to soil under lowland area and a resultant grain yield obtained in rice. The experiment was divided into two simultaneous studies, i.e., field experiment and metagenomic study, to obtain the rice yield and soil microbial dynamics, respectively. Pooled soil samples were collected from rice field with constant herbicide application, i.e. Pyraszosulfuron (pre-emergent) @25 g/ha + 2, 4-D (post-emergent @0.5 kg/ha), but with 5 different sources of nutrients, viz., T₀—absolute control, T₁—100% N-P₂O₅-K₂O through inorganic fertilizers (recommended dose of 40-20-20 kg/ha)), T₂—75% N through inorganic + 25% N through FYM (P₂O₅ and K₂O recommended doses), T₃—75% N through inorganic + 25% N through vermicompost (P₂O₅ and K₂O recommended doses) and T₄—75% N through inorganic + 25% N through crop residues and bio-fertilizer (P₂O₅ and K₂O recommended doses). Based on the amplicon DNA sequencing approach, it was observed that though there was an overall increase in bacterial phyla, viz., Chloroflexi, Actinomycetes, Euryarchaeota, Firmicutes in all the treatments from 0 days after transplanting (DAT) to physiological maturity of plant, however, soil treated with vermicompost (TH₃) showed a dramatic increase in the population of one particular microbe viz., Firmicutes. Amongst fungal populations, Actinomycetes increased in soils of all the treatments from 0 DAS to physiological maturity, however, the increase was lowest in soil treated with vermicompost while it was highest in soil having crop residues (T₄). The result obtained in microbial dynamics in case of vermicompost supplementation are concomitant to the biological yield of rice that was observed to be the highest in the same. The findings highlighted that the soil with vermicompost supplementation outperformed in terms of beneficial microbial changes and highest grain yield which, again, could be attributed to the favourable niche provided by vermicompost. Hence, supplementing vermicompost along with inorganic N sources can surpass the other organic sources in preventing deleterious effects of the chemical build-up in soil, due to herbicides and inorganic fertilizers, while synthesizing and releasing plant hormones, metabolites and antibiotics to suitably allow the growth and dominance of beneficial bacterial population.

Introduction

Soil representing one of the most diverse kinds of ecosystem on earth where the activities and interactions between different life forms explain about the health and productivity of soil. Soil acting as the hotspots to diverse conditions of bacteria, fungi, virus, protozoa etc., commonly referred to soil microbiota have been the source to the beneficial acids in the soil. The abundance of bacterial and fungal population in the soil is direct indicative of the increased soil health due to manipulated rate of biochemical cycles which are mostly correlated with differences that exist in soil chemistry and are discontinued or disrupted on slight changes in the soil physio-chemical properties^1,2. Not only the microbes are responsible for the degradation of organic remnants, but also humus formation, circulation of biogenic components along with their transformation into forms which are available for plants and also, degradation of pollutants³. It is the molecular substrates and the microclimate of the soil strata that, to a larger extent, determines increase or decrease in the microbial population. Availability of nutrients, changes in the chemical nature of soil, pesticide decomposition etc. are some other factors that have cumulative effects of the microbial population and diversity in soil. It is evident from the early researches that the chemical residues in soil disrupts the original structure of microbial population, thereby, evolving new biochemical cycles which may or may not favor the microbe-soil-plant relationship.

Rice suitably a semi-aquatic plant has been known to harbor diverse kinds of microorganisms owing to the oxic-anoxic zones that are formed under waterlogged condition. It has been reported in studies that microbial community structure in rice ecosystems is quite distinct from those observed in the other plants^4,5. Application of only inorganic sources of N, like urea, in waterlogging conditions are prone to losses such as volatilization, leaching, denitrification etc. and therefore supplementing organic manures such as VC, FY or CR would facilitate the microbial growth by providing them the substrate and hence holding the nutrients, especially N, for its slow release. The organic amendments increase the soil organic matter (SOM) and its decomposition is responsible for releasing various nutrients. However, under lowland rice conditions, the aerobic and anaerobic zonal distribution plays very important role in the distribution of the microbes and the kinetics of the microbial activities is very important to understand the rate of decomposition of various organic inputs. If the organic inputs take time to decompose, they create a labile carbon pool which may hinder the nutrient mobility. Under this condition as the very presence of the beneficial microbes decide the several other processes such as methane formation, ammonization, nitrogen fixation, nitrification etc. and hence, the correct nutrient sources addition is an important criterion in understanding their distribution.

Studies have evidently shown the correlation of the soil microbial communities, their mechanisms and roles in soil fertility with the incorporation of the organic or inorganic inputs into soil in due course of agricultural practices⁶. Although inorganic fertilizers are capable to provide available nutrients to plants in a shorter period of application, however, they are seldom responsible for contributing to soil health. Organic manures such as vermicompost, FYM, crop residues etc. can hence be supplemented along with inorganic fertilizers as an alternate source of N to fulfill the entire nitrogen requirements of the plants to boost microbial activities and improve soil health⁷. Organic manures act as buffer and are responsible for holding the nutrients through the action of chelation due to the presence of organic acids which are eventually released in the rhizosphere with the decomposition favored by the diverse microorganisms in soil. This causes further diversification of microbial population in soil which would either favor of some microbial class while being deleterious to the other. It is difficult to understand the complexity in rise and decline of microbial species in soil, especially on addition of organic manures, as this, to a larger extent, depend upon the type of substrate taken for compost preparation and the composting period.

Vermicompost is considered to comprise nutrient enriched peat-like substances as it undergoes proper decomposition of organic matter and microbial succession. FYM is a partially decomposed manure that harbors many aerobic and anaerobic microbes which releases nutrients upon decomposition in soil. They are good source of nitrogen, which is released in the rhizosphere, upon decomposition by the microbes⁸. Incorporating crop residues in soil has also been reported to favor microbial growth, improve soil organic matter and release nitrogen in soil upon mineralization⁹. These manures have been proven to effectively improve soil physical and chemical properties, soil respiration through improved enzymatic activities and therefore the fertility of soil.

Herbicide degradation under waterlogging condition is also a problem as they have low water solubility, low mobility and get adsorbed in the soil to persist for a longer duration. The kind of organic matter added to the soil to a large extent decides the rate of degradation of their residues, especially in the waterlogging condition where they may reach to a toxic level. Application of Pyrazosulfuron and 2,4-D as pre and post emergent herbicides, respectively, in rice is a common practice in major areas of northeast India which destroys the broad-spectrum weeds but at the same time it is known to be persistent in soil with a half-life of around 5–6 days in soil¹⁰and can have a considerable effect on microbial dynamics. The microbial activities and soil respiration effectively decreases in the rhizosphere as a result of the residual impact of these herbicides in soil^11,12,13.

Microbial diversity greatly influences the plant productivity^14,15, hence conceptualizing the microbial community in rice niche, as a result of different management practices is very crucial. The traditional cultural techniques are only capable of exploring 1% of the total microbial diversity present in soil¹⁶ which is insufficient to ascertain the actual changes in the microbial diversity. Metagenomic approach is one such technique that is culture-independent and gene-targeted which can be a step ahead in reaching to the rest 99% of the microbial diversities and quantifying their change. Several cultural and non-cultural studies have been conducted to read the microbial count in soil on single organic input supplementation, however very limited references are available to highlight the comparison in microbial alterations in soil as a result of their interactions with the various organic and inorganic chemicals in soil. Considering the aforementioned factors, the current research was conducted with the objectives: (a) to study the alterations in the microbial (bacterial and fungal) population, metagenomically, due to the different organic manures supplemented along inorganic fertilizers, to fulfill the 100% N requirement of crop, and herbicides in low-lying area and (b) to study the impact of the various N sources and herbicide applications on rice grain yield.

Methodology

Experimental site

An experiment was conducted for two years in Jorhat, Assam (26^o 48′ North latitude and 95^o 48′ East longitude) on rice-rice cropping sequence where summer rice, variety Luit, was grown from March to June and was followed by winter rice, variety Ranjit, grown from July to October. The area of field experiment was characterized by hot and humid summer followed by dry and cool winter with an average precipitation of around 2324 mm and average temperature of 24 ^oC.

Treatments details

The experiment was divided into two simultaneous studies, i.e. field experiment and metagenomic study in biotechnology laboratory in order to obtain the rice yield and soil microbial dynamics, respectively.

Feild Experiment

The field experiment was laid in Factorial Randomized Blocked design, replicated 3 times, for both the autumn and winter rice in sequence and consisted of 4 nutrient management treatments, viz., 100% Recommended dose of fertilizer (40:20:20 kg/ha N: P₂O₅:K₂O) (RDF); 75% Recommended dose of nitrogen + 25% N through Farm Yard Manure (RDF + FYM); 75% Recommended dose of nitrogen + 25% N through vermicompost (RDF + VC); and 75% Recommended dose of nitrogen + 25% N through Crop residues and Biofertilizers (RDF + CR); 3 weed management treatments, i.e., Pyrazosulfuron@25 g/ha + 2,4-D @0.5 kg/ha in both autumn and winter; Pyrazosulfuron@25 g/ha + 2,4-D @0.5 kg/ha in autumn rotated with Pretilachlor @0.75 kg/ha + 2,4-D @ 0.5 kg/ha in winter rice; and Pyrazosulfuron @25 g/ha + 2,4-D @0.5 kg/ha in autumn and only Pretilachlor@ 0.750 kg/ha in winter rice; and two controls, i.e., Farmer’s Practice (0.75 kg/ha pretilachlor + 20:10:10 kg/ha N: P₂O₅:K₂O) and Absolute control (no nutrient and weed management).

Metagenomic study

For metagenomic study, the soil samples (0–20 cm depth) were taken only during the second year of winter season which were from the plots with treatment combination of all the 4 nutrient management treatments given above with only 1 weed management treatment, i.e., Pyrazosulfuron@25 g/ha + 2,4-D @0.5 kg/ha in both autumn and winter; and 1 absolute control that totaled to 5 soil samples, each collected at two stages of rice, i.e., 0 DAT (Days after transplanting) and at physiological maturity. These samples collected were transferred to laboratory for metagenomic study.

Nutrient management in rice

Recommended dose of fertilizer for both the rice varieties was 40-20-20 kg/ha N, P₂O₄ and K₂O, where, nitrogen was given as per the different treatment requirements and entire phosphorus and potassium was given initially at the time of sowing. The organic manures, i.e., VC, FYM and CR were applied to the respective treatment plots to supplement 25% of total nitrogen requirements of rice, whereas, rest 75% was supplied through inorganic fertilizers, i.e., Urea, SSP and MOP. Both vermicompost (VC) and farm yard manure (FYM) was prepared using chopped rice straw and cowdung as the primary raw materials using bed method and pit method, respectively. Well decomposed VC was harvested in 28 days, whereas, FYM was harvested in 4 months after 2–3 rounds of turning over for proper aeration. Post harvesting, 5 g of VC and FYM were sampled for nutrient analysis. Nitrogen analysis was done using alkaline potassium permanganate¹⁷method. On dry weight basis, VC was observed to contain 1.5% N and FYM consisted 0.43% N. Similarly, crop residues (CR) of previously harvested rice was used to determine the N content through Kjeldahl digestion method¹⁸ and was observed to contain 0.56% N content. Accordingly, rice straw @5 kg/ha was incorporated into the soil of treatment plots, 10 days before transplanting of rice seedlings, after spraying with the cellulose decomposing bacteria@ 5 kg/ha. Azotobacter@ 3 kg/ha was applied to the same fields after incorporating crop residues, considering azotobacter fixes 25–30 kg N/ha in soil.

Statistical analysis of field data

Statistical analysis was done for yield data of rice using Excel 2021 software while pooled data analysis for ANOVA was performed using SPSS 19. The statistical significance of various effects was tested at 5% probability level. Two tailed t-test (at alpha 0.05) was done for comparing the nutrient and weed management treatments’ means with the control (Absolute control and farmer’s practice) means.

Study area and sample collection for metagenomic studies

For metagenomic study, soil samples were taken from the rice rhizosphere only during the winter season of last year. Treatment combinations of all the 4 nutrient management treatments were taken with only 1 weed management treatment, i.e., Pyrazosulfuron@25 g/ha + 2,4-D @0.5 kg/ha in both autumn and winter; and 1 sample was drawn from Absolute control. A total of 10 samples were collected at two different stages of plant growth, i.e., 5 samples at 0 DAT of rice and 5 samples at physiological maturity of rice. Soil sample (~ 500 g) was collected and pooled from 5 random locations of each plot 0–20 cm depth of soil and was repeated for all the three replications. Again, the soil from different replications of similar treatments were pooled to make it a common sample of one. Hence, we pooled 5 samples in total, during both the years, each at 0 DAT (day after transplanting) and at physiological maturity of crop and they were labelled as mentioned in Table 1.

Table 1.

Details of the collected soil samples for metagenomic study.

Samples taken at 0 DAT of winter rice		Samples at physiological maturity (120 DAT) of winter rice
T0	Absolute control (without nutrients and weed management)	TH0	Absolute control (without nutrients and weed management)
T1	100% N-P2O5-K2O through inorganic fertilizers (recommended dose of 40-20-20 kg/ha)) + Pyrazosulfuron @0.025 kg/ha + 2,4-D @ 0.5 kg/ha	TH1	100% N-P2O5-K2O through inorganic fertilizers (recommended dose of 40-20-20 kg/ha)) + Pyrazosulfuron @0.025 kg/ha + 2,4-D 0.5 kg/ha
T2	75% N through inorganic + 25% N through FYM (P2O5 and K2O recommended doses) + Pyrazosulfuron @0.025 kg/ha + 2,4-D @0.5 kg/ha	TH2	75% N through inorganic + 25% N through FYM (P2O5 and K2O recommended doses) + Pyrazosulfuron @0.025 kg/ha + 2,4-D @0.5 kg/ha
T3	75% N through inorganic + 25% N through vermicompost (P2O5 and K2O recommended doses) + Pyrazosulfuron @0.025 kg/ha + 2,4-D @0.5 kg/ha	TH3	75% N through inorganic + 25% N through vermicompost (P2O5 and K2O recommended doses) + Pyrazosulfuron @0.025 kg/ha + 2,4-D @0.5 kg/ha
T4	75% N through inorganic + 25% N through Crop residues and bio-fertilizer (P2O5 and K2O recommended doses) + Pyrazosulfuron @0.025 kg/ha + 2,4-D @0.5 kg/ha	TH4	75% N through inorganic + 25% N through Crop residues and bio-fertilizer (P2O5 and K2O recommended doses) + Pyrazosulfuron @0.025 kg/ha + 2,4-D @0.5 kg/ha

Metagenomic DNA isolation

Metagenomic DNA was extracted from the soil samples using the commercial soil extraction kit (Nucleospin Soil, Mobio, USA) as per manufacture’s instruction. The quality and quantity of the isolated genomic DNA samples was checked using NanoDrop (Thermo, USA) and Qubit Fluorimeter (V.3.0). The high-quality DNA samples were thereafter processed for amplicon generation which led to the generation of NGS library.

Preparation of 16S rRNA and 18s rRNA amplicon-based HiSeq library

The V3–V4 region of 16S rRNA gene was amplified from the metagenomic DNA of each sample using specific V3 Forward primer CCTACGGGNBGCASCAG and V4 Reverse primer GACTACNVGGGTATCTAATCC. Similarly, The ITS1-ITS2 Inter-transcribed region was amplified using specific ITS1 Forward primer TCCGTAGGTGAACCTGCGG and ITS2 Reverse primer GCTGCGTTCTTCATCGATGC. The PCR reaction for amplification of V3-V4 and ITS1-ITS2 Inter-transcribed region was performed separately in a total volume of 25 µl containing 2.5 µl of metagenomic DNA, 1 µl of each primer (10 µM), and 12.5 µl of 2× ReadyMix (Takara, Japan). The PCR was carried out using the thermal cycler program: 95 °C for 3 min followed by 25 cycles at 95 °C for 30 s, 55 °C for 30 s, 72 °C for 30 s and a final extension at 72 °C for 5 min with a holding at 4 °C. The amplified products were checked on 2% agarose gel and gel purification were done to remove non-specific amplifications. The amplified product (5ng) was used for library preparation using NEBNext Ultra DNA library preparation kit as per manufacture’s instruction (NEB). The library quantification and quality estimation were carried out in an Agilent 2200 Tape Station. The prepared pair end libraries (2 × 250 bp) were then sequenced in an Illumina HiSeq 2500 platform.

Analysis of metagenomics data

The preprocessed V3-V4 and ITS1-ITS2 consensus sequences were analyzed using QIIME (version 2.0) pipeline. QIIME is a comprehensive software comprising tools and algorithms to explore phylogenetic inferences and assignment of taxonomic data using naïve Bayesian classifier. Pre-processed reads from all samples were pooled and clustered into Operational Taxonomic Units (OTUs) based on their sequence similarity using U cluster program (similarity cutoff. = 0.97) available in QIIME software and the OTUs with greater than 5 reads were considered. The representative sequences from each clustered OTUs were picked and aligned against SILVA core set of sequences using Py NAST program. Further, taxonomy classification was performed using RDP classifier by mapping each representative sequence against SILVA OTUs database. Microbial community composition and relative abundance profiles of OTUs of each samples taxonomic distribution at phylum level were analyzed. It should be noted that the taxa other than top 10 are categorized as ‘Others’ while the sequences that does not have any alignment against taxonomic database are categorized as ‘Unknown’.

Alpha diversity and rarefaction curve were analyzed to calculate species richness. The microbial diversity within the samples was analyzed by calculating Shannon, Chao1 and observed species metrics. The chao1 metric estimates the species richness while Shannon metric is the measure to estimate the observed OTU abundances, and accounts for both richness and evenness. The observed species metric is the count of unique OTUs identified in the sample. The metric calculation was performed using QIIME software.

Comparison of microbial communities between the samples was also performed. At first, the distance matrix was generated using both weighted and unweighted UniFrac approach. Sequence abundances were taken in to account in Weighted UniFrac for comparing microbial diversity. A jackknife test was performed to construct a consensus UPGMA (Unweighted Pair Group Method with Arithmetic Mean) tree for all samples collected at 30 days and 60 days separately. The consensus sequences were taken to construct UPGMA trees by using weighted and unweighted UniFrac distance matrix.

Results

Metagenomic study

With the aforementioned analysis, the composition of bacterial and fungal population was compared in two ways, i.e., comparison of each between the two sampled stages- 0 DAT and physiological maturity; and comparison between the 5 treatments at each sampled stage.

1. Comparison of bacterial diversity at 0 DAT and physiological maturity.

   Based on the V3-V4 sequencing results of the samples it was observed that bacterial community dominated the soil microbial composition. The OTUs percentage of the bacterial phyla of absolute control (Pooled Sample-1) was observed for the both the days i.e., 0 DAT (T₀) and at physiological maturity (TH₀). Among the bacterial phylum, Proteobacteria was the most abundant of the total bacterial population followed by Cloroflexi, Acidobacteria, and Others in T₀ as well as TH₀ (Fig. 1a). The phyla Bacteroidetes, Verrucomicrobia, Firmicutes and Actinobacteria were also observed in both the samples. Bacteroidetes, Acidobacteria, Verrucomicrobia and Others were higher in T₀ while those of Actinobacteria, Chloroflexi and Firmicutes were higher in TH₀.

   OTU percentage, for the treatment 100% N applied from RDF along with Pyrazosulfuron @25 g/ha + 2,4-D (Pooled Sample-2), was observed in both the samples that were collected at 0 DAT (T₁) and physiological maturity (TH₁). Proteobacteria was found to be the most abundant bacterial phyla followed by Acidobacteria, Cloroflexi, Others and Bacteroidetes, in addition of small fractions of Euryarchaeota and Nanoarchaeota in both the samples. However, at physiological maturity (TH₁) the proportion of all the phylum reduced except those of Euryarchaeota, Others and Actinobacteria. (Fig. 1b).

   Soil samples of the treatment- 75% N from RDF + 25% N from FYM along with Pyrazosulfuron @25 g/ha + 2,4-D (Pooled Sample-3), were also dominated by the phylum Proteobacteria which was followed by Acidobacteria, Cloroflexi, and Others. Bacteroidetes and Verrucomicrobia were also detected in sufficient amount. However, Proteobacteria, Acidobacteria, Bacteroidetes, and Verrucomicrobia reduced at physiological maturity (TH₂) while those of Euryarchaeota, Actinobacteria and Firmicutes increased. (Fig. 1c).

   Soil samples of the treatment- 75% N from RDF + 25% N from vermicompost along with Pyrazosulfuron @25 g/ha + 2,4-D (Pooled Sample-4), that were collected at 0 DAT (T₃) and physiological maturity (TH₃) revealed dominance of Proteobacteria followed by Cloroflexi, Others and Bacteroidetes. The content of Proteobacteria, Bacteroidetes and Verrucomicrobia reduced at physiological maturity (TH₃), whereas, the content of Euryarchaeota, Actinobacteria, Chloroflexi and Firmicutes increased. (Fig. 1d).

   Similarly, the samples of the treatment- 75% N from RDF + 25% N from crop residues along with Pyrazosulfuron @25 g/ha + 2,4-D (Pooled Sample-5), collected at 0 DAT (T₄) and physiological maturity (TH₄) showed higher abundance of Proteobacteria followed by Chloroflexi, others and Acidobacteria. Proportion of Proteobacteria, Acidobacteria, Bacteroidetes, Verrucomicrobia and Firmicutes reduced at physiological maturity while those of phylum Chloroflexi, Others, Actinobacteria, Euryarchaeota and unknown were higher. (Fig. 1e). There was no change in the proportion of Firmicutes.

2. Inter-treatment comparison of bacterial diversity at 0 DAT.

   V3-V4 sequence data of the soil samples collected at 0 DAT revealed that the phylum Proteobacteria dominated all the 5 samples i.e. T₀, T₁, T₂, T₃ and T₄, and its abundance was highest in the sample T₂. They did not vary and remained unchanged as there were no changes in the bacterial composition due to treatments. Also, phylum Proteobacteria dominated in all the treatments. (Fig. 2a).

3. Inter-treatment comparison of bacterial diversity at physiological maturity.

   On comparing all the treatments at physiological maturity (Fig. 2b), a similar pattern of increase in the populations of Actinobacteria, Cloroflexi, Acidobacteria, Verrucomicrobia, Euryarchaeota, Unknown and Others were found in all the treatments, i.e. TH₀, TH₁, TH₂, TH₃ and TH₄. However, on comparing all the treatments, phylum Firmicutes were observed to increase dramatically at physiological maturity of crop. Though the population of Firmicutes increased in all the treatments at physiological maturity, they were found to be highest in the treatment, TH₃, i.e., vermicompost along with inorganic nutrient application.

4. Comparison of fungal diversity between samples of 0 DAT and physiological maturity.

   Analysis of the OTUs obtained from the ITS1-ITS2 sequenced data revealed that the studied samples harbored different phyla of fungi. Analysis of OTUs of samples of both the stages for absolute control, i.e., 0 DAT (T₀) and physiological maturity (TH₀), treatment- 100% N-P₂O₅-K₂O through inorganic fertilizers along with Pyrazosulfuron @25 g/ ha + 2,4-D, that were collected at 0 DAT (T₁) and physiological maturity (TH₁) and treatment- 75% N from RDF + 25% N from FYM along with Pyrazosulfuron @25 g/ha + 2,4-D collected at 0 DAT (T₂) and physiological maturity (TH₂) showed that the phylum Ascomycota, Basidiomycota and Unknown dominated the soil. However, the abundance of Ascomycota and Basidiomycota increased at the physiological maturity (TH₀, TH₁ and TH₂) while those of ‘Unknown’ reduced. ‘Others’ were also detected at the physiological maturity in the treatment with 100% inorganic fertilizer (TH₁) (Fig. 3a, b and c).

   Soil sample of the treatment- 75% N from RDF + 25% N from vermicompost along with Pyrazosulfuron @25 g/ha + 2,4-D, collected at 0 DAT (T₃) and physiological maturity (TH₃) were also dominated by the phyla Ascomycota, Unknown, and Basidiomycota. However, the content of Ascomycota, and Basidiomycota were higher at 0 DAT (T₃) while they remain unchanged at physiological maturity. Abundance of ‘Unknown’ was observed to be higher at physiological maturity (TH₃) (Fig. 3d).

   In the treatment- 75% N from RDF + 25% N from crop residues along with Pyrazosulfuron @25 g/ha + 2,4-D, abundance of Ascomycota increased at physiological maturity (TH₄) and abundance of ‘Unknown’ decreased (Fig. 3e).

5. Inter- treatment comparison of fungal diversity in samples collected at 0 DAT.

   Similarly, the comparison between the samples, on the basis of ITS1-ITS2 sequencing, of different treatments that were collected at 0 DAT, i.e., T₀₋, T₁, T₂, T₃ and T_4, revealed DAT showed that all the samples were dominated by Unknown phyla followed by Ascomycota and Basidiomycota (Fig. 4a).

6. Inter- treatment comparison of fungal diversity in samples collected at physiological maturity.

   At physiological maturity, a small fraction the phylum Zygomycota was detected in the sample of TH₁ and TH₀. The abundance of Ascomycota was highest in the sample TH₄ that was followed by the samples TH₂ and TH₁. The population of Basidiomycetes remained high in all the treatments at physiological maturity except in the sample TH₃ (Fig. 4b).

7. Alpha diversity.

   The species richness was calculated from the individual samples on the basis of ‘rarefaction curves where the Shannon metric is the measure to estimate observed OTU abundances, and accounts for both richness and evenness. The curve represented the plot of total numbers of distinct species which is annotated as function of number of sequenced samples. Presence of the steep slope on the left indicated presence of a larger fraction of species diversity which is yet to be discovered. Vertical axis represents the diversity of the community, while horizontal axis represents the number of sequences which are considered in diversity calculation. The refraction curve analysis of the V3-V4 sequence data of 0 DAT sample showed that the sample T₃ had the highest observed bacterial species followed by T₀, T₁, T₂ and T₄ respectively. However, in the V3-V4 sequence data of physiological maturity samples, TH₁ recorded the highest observed bacterial species followed by TH₀, TH₂, TH₃ and TH_4, respectively. The refraction curves of the ITS1-ITS2 sequence data of the 0 DAT samples showed that the sample T₂ had the highest observed fungal species followed by T₀, T₃, T₄ and T₁ respectively while in the ITS1-ITS2 sequence data of physiological maturity samples, TH₃ recorded the highest observed fungal species followed by TH₀, TH₂, TH₁ and TH₄ respectively. Thus, the samples collected were optimum in obtaining the species richness of bacterial and fungal phyla in the niche.

   The refraction curves prepared based on Shannon for bacterial and fungal phyla are represented in Figs. 5 and 6 respectively.

8. Beta diversity.

   Comparison of microbes between the samples of the 16 S and ITS sequence data of the 0 DAT samples was performed using weighted ‘UniFrac approach’ that represented the presence of microbes and was represented through Unweighted Pair Group Method with Arithmetic Mean (UPGMA) tree. Cluster tree prepared from V3-V4 sequences of the samples at 0 DAT samples which are based on weighted unifrac approach indicated that the bacterial diversity of the sample T₁ and T₃ as well as T₀ and T₄ are similar but the bacterial diversity of T₂ is different from the other samples. While the cluster tree of the ITS sequence based on weighted unifrac approach indicated that fungal diversity of the sample T₃ and T₄ is highly similar whereas the fungal diversity of T₀, T₁ and T₂ are different.

   Cluster tree prepared from the V3-V4 sequences of the samples at physiological maturity, based on weighted unifrac approach revealed that the bacterial diversity of the sample TH₀ and TH₁ as well as TH₂ and TH₄ are similar but the bacterial diversity of the sample TH₃ differs from the other samples. The Cluster tree of the ITS sequence data of the samples at physiological maturity, based on weighted ‘Unifrac approach’ revealed that fungal diversity of the sample TH₂ and TH₃ is highly similar whereas those of TH₄, TH₀ and TH₁ are different. The Cluster trees representing weighted Unifrac approach is shown in Figs. 7 and 8.

Fig. 1.

Comparing bacterial OTU reads of 0 DAT and physiological maturity at Phylum level.

Fig. 2.

Comparison of bacterial diversity at phylum level (OTUs) between 0 DAT and at physiological maturity.

Fig. 3.

Comparing fungal OTU reads of 0 DAT and physiological maturity at Phylum level.

Fig. 4.

Comparison of fungal diversity at phylum level (OTUs) between 0 DAT and at physiological maturity.

Fig. 5.

Rarefaction curves showing no. of individuals for bacterial diversity on x-axis and no. of species on y-axis.

Fig. 6.

Rarefaction curves showing no. of individuals for fungal diversity on x-axis and no. of species on y-axis.

Fig. 7.

Beta diversity analysis between samples for bacterial population.

Fig. 8.

Beta diversity analysis between samples for fungal population.

Grain yield (kg/ha)

Grain yield as computed was found to be significantly higher in the treatment, RDF + VC as compared to the other nutrient management treatments that was followed by RDF + FYM during both the years of autumn and winter rice (Table 2). Grain yield was statistically analyzed individually for different weed managements also and it was observed that among the weed management treatments, W₂- Pyr. + 2,4-D rotated with Pret. + 2,4-D, resulted in significantly higher grain yield of rice. However, to study non-rotated herbicide effects on microbes, in our metagenomic assay, we had considered the only treatment, W₁- Pyrazosulfuron@25 g/ha + 2,4-D @0.5 kg/ha in both autumn and winter. The two tailed t-test applied for comparing the treatments with control suggested that nutrient and weed management treatments were significantly higher than both the controls, i.e., C₀ and C₁, in both the autumn and winter rice during both the years.

Table 2.

Grain yield(kg/ha) recorded with various nutrient and weed management treatments.

Grain yield (q/ha)
	Autumn rice 2018	Autumn rice 2019	Winter rice 2018	Winter rice 2019
Nutrient management (T)
T1- 100% RDN	30.98	31.22	35.05	35.08
T2- 75% RDN + 25% FYM	34.08	34.41	36.13	36.40
T3- 75% RDN + 25% VC	36.27	36.37	37.70	37.35
T4- 75% RDN + 25% CR	32.85	33.85	35.50	35.71
S.Ed (±)	0.53	0. 51	0.49	0.41
CD (t at 5%)	1.10	1.06	1.02	0.86
Weed management (W)
W1- Pyr. +2,4-D	33.07	33.74	35.65	36.07
W2- Pyr. + 2,4-D rotated with Pret. + 2,4-D	35.17	35.45	37.66	37.30
W3- Pyr. +2,4-D in autumn; Pret. in winter	32.40	32.69	34.97	35.03
S.Ed (±)	0.46	0.44	0.42	0.36
CD (t at 5%)	0.96	0.92	0.88	0.75
(TxW)	1.91	1.84	1.76	1.49
Control
C0 - Absolute control	28.76	28.14	30.91	31.17
C1 - Pret. (20-10-10 kg/ha N-P205 -K2 O)	29.65	30.36	33.69	3354
Two tailed t-test (t at 5%)
T vs. C0	0.023	0.007	0.002	0.004
T vs. C1	0.041	0.047	0.014	0.008
W vs. C0	0.012	0.002	0.003	0.012
W vs. C1	0.017	0.016	0.014	0.021

Discussion

Metagenomic study

In all the treatments, except the one with vermicompost (TH₃), Proteobacteria dominated at physiological maturity that was followed by Cloroflexi, Bacteroidetes and Acidobacteria. In one of the studies was found that around 73–86% of the total sequences studied from tropical agricultural land was dominated by Acidobacteria, Proteobacteria, Firmicutes, Actinobacteria, Verrucomicrobia, Gemmatimonadetes and Bacteroidetes^19,20. Similary, Proteobacteria, Acidobacteria, Actinobacteria and Bacteroideteshave been discovered as the dominant phyla in arctic soils²¹. Scientists also found that Proteobacteria, Acidobacteria, Actinobacteria and Betaproteobacteriawere the dominant taxa in German forest soils²²and that they are commonly found in soil environments as obligate or facultative anaerobes and are related to a wide range of functions including nitrogen, carbon and Sulphur cycling producing phytohormones which are crucial for the plant growth. They naturally occupy the highest richness in the soil^23,24 and are known to manipulate the population of other bacterial and fungal pathogens due to enzyme and antibiotic synthesis. In the soil supplemented with vermicompost (RDF + VC), however, the dramatic increase in Firmicutes population was observed at physiological maturity. Firmicutes are the group of bacteria that are widely known for the plant growth promotion, suppressing the plant pathogens and phytoremediation of pesticides and heavy metals. They are resistant to extreme surroundings and have the capability to reproduce in under toxic niche because of the Gram-positive cell wall they possess.

Firmicutes, though, found to increase in all the treatments, however it majorly increased in the treatment where vermicompost was added along with inorganic fertilizers (RDF + VC) that can be correlated to increased nutrient availability, mobility of nutrients, presence of plant hormones and enzymes in soil. The dominance of this bacteria is also indicative of its ability to degrade the herbicide residues better as compared to the other treatments. Aresearch reported similar observation, wherein it was concluded that the members of phylum, Actinobacteria, Firmicutes and Proteobacteriawere responsible for disintegration of pesticides of diverse groups in soil²⁵ and hence, there population increased in soil because of this fact. The increase in Firmicutescontent was also observed in forest soil of Kashmir, India²⁶, which was correlated with increase in nutrient like phosphorus and optimizing soil pH. Vermicompost acts as buffer that normalizes the pH due to its nutrient retention properties. Because it is well decomposed, the time taken by the nutrients to get available to plants is relatively lesser as compared to FYM or crop residues. Research also reported similar case, where it was confirmed that increase in the organic input in soil constructed soil fertility and as a result showed increased abundance of Firmicutes^27,28.

Phylum Ascomycota and Basidiomycota dominate the fungal phyla in any soil that account for more than 70% of fungal species and they mostly constitute the plant pathogenic subphyla like Taphrinomycotina, Saccharomycotina and Pezizomycotina (Ascomycota); and Pucciniomycotina, Ustilaginomycotina and Pucciniomycotina (Basidiomycota). They can survive in permanently waterlogged and poorly aerated soil, such as submerged paddy soils²⁹are more important than bacteria in decomposing soil organic matter (SOM), especially in acidic ecosystems^30,31. They represent the main classical fungal decomposers in different soils³² being saprophytic in nature. Their abundance in the later stages might be because of their extremophilic nature irrespective of the soil organic horizons, explaining why Ascomycota and Basidiomycota outperformed as the dominant species in TH₁, TH₂ and TH₄^33,34. The study is similar under the microbial straw decomposition^35,36. Fungal populations tend to dominate in the soil with lower acidity, low-nutrient status with recalcitrant litter ad high C: N ratios³⁷. The increased population of both the phyla can also be reasoned by their ability of decomposing the herbicides better as compared to other fungal phyla. Similar findings are reported³⁸, wherein wood rot fungi group, belonging to Ascomycota and Basidiomycota were proven to be capable of degrading a broad range of polycylic aromatic compounds, phenols, chlorinated biphenyls and other compounds.

The reduced abundance of phyla Ascomycota and Basidiomycota in TH₃ can be attributed to increased abundance of Firmicutes in the treatment and the bacterial-fungal competition in the rhizosphere. Firmicutes consist of species like Clostridium, Peptococcus, Bacillus, Mycoplasm, Symbiobacterium etc. some of which possess endospore formations and hence can dominate in the anaerobic soil conditions with ease. Increase in the bacterial population as favored by vermicompost addition might have led to overcome the competition among bacterial and fungal species for the same substrate, which led to the reduction in population of the saprophytic fungal phyla in TH₃^39,40. Dominance of Firmicutes in soil also indicates ability of the gram-positive bacteria to have negative correlation with the fungus like Ascomycota and Basidiomycota⁴¹. Similar anti-fungal properties of firmicutes owing to their resistant nature, and as a result suppression of fungal population, has been reported in researches^{42,43,44,45,46,47}.

Although the abundance of Basidiomycota increased in later stages, it was relatively lower than Ascomycota. Basidiomycotaconsist wide range of species that are mycorrhizal or saprophytic. The lower abundance of the phyla can be attributed to the fact that they are adversely affected by the N-fertilization, particularly being the taxa that encompass mycorrhizal fungi classes⁴⁸.

Grain yield

The significantly higher grain yield in treatment supplemented with vermicompost (RDF + VC) could be the result of the nutrient availability in the rhizosphere that can be correlated with the favorable growth of bacteria, viz., Firmicutes and Proteobacteria and the suppression of fungus like Ascomycota and Basidiomycota. The result indicates that addition of VC to soil allows the growth of beneficial bacteria in the waterlogging condition of rice while irrespective of the deleterious effect of the herbicide in soil, resulting in significantly higher grain yield as compared to the other nutrient management treatments.

Conclusion

Due to the oxi-anoxic rhizospheral environment, bacterial and fungal composition in rice rhizosphere is notably disparate from the rest of the crops⁴⁹ creating a separate form of niches for the biochemical reactions in soil on addition if different inputs. In this research we had critically analyzed the effect of different organic sources, i.e., VC, FYM and CR, that were supplemented along with inorganic fertilizers and herbicides and it was found that VC allowed the beneficial microbial alterations and higher grain yield. Addition of vermicompost to soil could be effective in enhancing the soil health due to the favorable changes as a result of protruding bacterial population such as Proteobacteria and Firmicutes while decreased population of fungal population viz., Ascomycota and Basidiomycota. This gives the scope of improvement in the soil productivity, soil organic matter and richness in the useful soil microbes that can be favorable under the waterlogging niches found in lowland rice cultivations. The impact of vermicompost addition in soil was concomitant to the increased yield of rice in the treatment as compared to other treatments, during both the years of autumn and winter season. Hence, VC can be considered as a better source of N that can be supplemented along with inorganic fertilizers and herbicides as it would help in the richness of good bacteria thereby increasing yield in waterlogging scenarios. Further researches are required to study the impact of different organic and inorganic sources of nutrients on microbes and crop productivity under different ecological scenarios and to find out the best suitable management practice to augment the beneficial microbial population.

Author contributions

All the authors share equal contributions to the paper. Collection and analysis of data has been done by the corresponding author. The drafting of portion which consist introduction, materials and methodology has been done by the first author while drafting of the result and discussions is done by the third author.

Funding

The work has been supported internally by Assam Agricultural University, Jorhat, Assam, where the work was carried out, under the research proposals grant.

Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.