Filtered mud improves sugarcane growth and modifies the functional abundance and structure of soil microbial populations

Soil substrate preparation and experimental design

The soil used for the experiment was collected from the Fujian Agriculture and Forestry University sugarcane cultivation field (latitude: 26°05 9.60″N; longitude: 119°14 3.60″E). The physicochemical properties of soil were; total carbon (TC): 0.97%, total phosphorus (TP): 0.7 g/kg, and total nitrogen (TN): 0.10%. The soil particle size distribution was clay: 20.3% silt: 43.1% sand: 36.6%, which was classified as clay-loam. The sugar mill filtered mud (FM) used in this experiment was obtained from Nanjing Qinfeng Crop Straw Technology Company, China. Before mixing with the soil, the FM was oven-dried at 50 °C until a constant weight was achieved. The chemical properties of the FM are as follows; pH: 7.28, EC: 3.45 dS/m, O.C: 23.31%, O.M: 40.1%, N: 1.64%, P: 14.4 g/kg, K: 14.1 g/kg, Ca: 1.33 cmol/kg, Mg: 1.37 cmol/kg, S: 1.52%, and Na: 1.44 cmol/kg.

The experiment was carried out in a greenhouse at the Fujian Agriculture and Forestry University Fuzhou, Fujian Province, P.R China, from March to December 2019 using red PVC pots. Each pot had a height of 180 mm and a diameter of 120 mm. The experimental soil was properly air-dried and sieved into a 2 mm size fraction and mixed according to the respective ratios with FM to give a total weight of 10 kg per pot. The treatments evaluated comprised of a control (CK) without any amendment and three different FM:soil ratios. These FM ratios were hypothetically selected to accommodate its possible proportions that can be used under various conditions, from fields requiring more soil to greenhouse cultivation requiring less soil. These treatments were: CK, FM1 (FM:soil at 1:4), FM2 (FM:soil at 2:3), and FM3 (FM:soil at 3:2). The treatments were replicated three times and arranged in a randomized complete block design.

After mixing the soil and FM according to the respective proportions, the pots were watered to field capacity and left for 24 h to equilibrate. Sugarcane stems were planted vertically in each pot and periodically irrigated to field capacity using tap water. Each pot was also kept weed-free by hand weeding throughout the experimental period. During the experiment, the air temperature recorded was 25–27 °C, and relative humidity of 75–80% was maintained throughout the experiment. The agronomic characteristics (as listed in the plant data collection section) of the sugarcane plants were recorded at the tillering stage, which signified the period of rapid nutrient uptake for reproductive growth.

At the plant’s tillering stage (90 days after planting), about 40 g of soil sample was collected from each pot’s root zone using a portable soil auger of 1 cm diameter. Six subsamples were taken around and within the root zone of the crop per pot up to a depth of 18 cm to represent the rhizosphere soil and bulked to form a composite. A subsample of fresh soil was taken and put in well-labeled bags and stored in a refrigerator at 4 °C for the evaluation of the soil enzyme activity. Part of the subsampled soil was air-dried, ground, and sieved using a 2-mm sieve to analyze the soil’s physicochemical properties. About 10 g of the fresh soil subsamples were collected and stored at −20 °C for DNA extraction.

Chemical properties of the soil

The pH of the soil samples was measured with a glass electrode pH meter using 1:2.5 (weight/volume) soil:water ratio (Ibrahim et al., 2020b). Electrical conductivity (EC) was measured with a conductivity meter, while the soil total C and N were measured using the Flash Smart elemental analyzer (Thermo Scientific™, Waltham, MA, USA). The soil available K was determined by extraction using the ammonium acetate method and measured using flame photometry (Pansu & Gautheyrou, 2007). The molybdenum blue method was used to measure the available phosphorus (AP). Soil ammonium (NH₄⁺) and nitrate (NO₃⁻) were extracted using 2 M KCl, and measured by a Bran+Luebbe GmbH-AutoAnalyzer 3 (Bran+Luebbe, Norderstedt, Germany). The activities of the soil urease, phosphatase, β-Glucosidase, and cellulase were measured as described by Sun et al. (2014). Soil urease activity assay was based on the NH₄⁺−N released when the samples were incubated with 10 mL of 10% urea solution and 20 mL of citric acid buffer (pH 6.7) at 37 °C for 24 h. Cellulase activity assay involved the determination of reducing sugars produced when the soil sample was incubated with acetate buffer (50 mM, pH 5.5), carboxymethyl cellulose, and toluene at 37 °C for 24 h (Deng & Tabatabai, 1994). Soil phosphatase activity was determined after incubating the samples with 0.25 mL of toluene and 1 mL of disodium p-nitrophenyl phosphate tetrahydrate and placed in a water bath for 1 h at 37 °C (Tabatabai & Bremner, 1969). The assay for the determination of the β-glucosidase enzyme activity was carried out after incubating the samples with 50 mM cellobiose substrate solution in citrate–phosphate buffer (pH 6.30) in a shaker at 37 °C for 1 h (Stege et al., 2010).

Plant data collection

The plants’ height (PLH) and diameter (PLD) were measured with the aid of a meter rule and Vernier caliper, respectively. The PLH in each pot was measured from the base of the plant to the top visible dewlap (TVD) leaf. The Vanier caliper was used to measure the stalk diameter around the middle of the stalk. The number of leaves (NL) and the number of tillers (NT) on each plant were counted and averaged at the tillering stage. In addition, the leaves’ chlorophyll content (Chol) was measured using a chlorophyll meter (SPAD-502; Minolta, Osaka, Japan).

Soil genomic DNA extraction and amplicon sequencing

The total genomic DNA of the soil was extracted from 0.5 g of each soil sample using the Fast DNA^TM Spin kit for soil following the manufacturer’s instructions (MP Biomedical, Solon, OH, USA). The extracted soil DNA’s quality was visualized on 1% gel electrophoresis and assessed using a NanoDrop ND-2000 spectrophotometer (NanoDrop Technologies, Thermo Scientific, Wilmington, DE, USA). The DNA concentration was thereafter quantified using the Qubit assay (Invitrogen, CA, USA).

The amplification of the v3–v4 region of the 16S rRNA genes in the extracted DNA as well as the v4 region of 18S rRNA were carried out using the polymerase chain reaction (PCR). The 341F/797R forward and reverse primers with barcodes were used for the PCR amplification of the 16S rRNA genes (Caporaso et al., 2010), while the 1176F/1536R were used for that of the 18S rRNA genes (Smit et al., 1999). The PCR amplification was carried out on each samples using three replicates, with a mixture containing 4 μL of 5 × Reaction Buffer, 2 μL of dNTPs (2.5 mM), 0.25 μL of TransStart Fastpfu DNA Polymerase, 1 μL of each primer (5 μM) (reverse and forward), and ~10 ng template DNA (Michelsen et al., 2014). The PCR conditions involved an initial denaturation of 98 °C for 1 min, followed by 30 cycles of denaturation at 98 °C for 10 s, annealing at 50 °C for 30 s, extension at 72 °C for 60 s, and a final extension at 72 °C for 5 min. A 2% agarose gel electrophoresis was carried out to visualize the amplified DNA. The amplified PCR products were thereafter recovered by cutting the gel using the AxyPrep DNA Gel Recovery Kit (Axygen Biosciences, CA, USA). Samples that are characterized by one bright main strip between 400 bp and 450 bp were selected for further sequencing. The purification of the obtained PCR product was subsequently done using a Gel Extraction kit (TIANGEN, Beijing, China). An equal concentration of the purified amplicons was mixed into a single tube for sequencing. The initiation of the sequencing libraries was carried out in Illumina with the aid of a specific NEB Next® Ultra™ DNA Library Prep Kit. The library sequencing quality having an average insert size of 400 bp was used for paired-end sequencing on an Illumina MiSeq platform at the New England Biolabs Ltd. Beijing, China. The raw reads generated from amplicon sequencing were submitted to the Sequence Read Archive of the NCBI (https://trace.ncbi.nlm.nih.gov/Traces/sra/) under the bioProject PRJNA755433, and biosample accessions from SAMN20822645 to SAMN20822648 for soil bacteria, and SAMN20822649 to SAMN20822652 for the fungi.

Statistical analysis and bioinformatics

Data obtained from plant characteristics and soil physicochemical properties were analyzed using the analysis of variance (ANOVA), and the means obtained were separated using Tukey’s test at p < 0.05.

Quantitative Insights into Microbial Ecology (QIIME2; https://qiime2.org) pipeline (Bolyen et al., 2019) was used to classify raw sequences according to the specific barcode assigned to each sample. The original DNA segments were combined with the Paired-end reads (PERs), using FLASH (Baltimore, MD, USA) (Magoč & Salzberg, 2011). PERs were designed for each sample according to the particular barcodes connected with DNA fragments. UPARSE-OTU reference algorithm was used to analyze the sequences with the UPARSE software package (CA, USA). Truncation and removal of chimeric sequences were done using the UCHIME algorithm. It was ensured that the average quality score remained above 33 with regards to the 50 bp sliding window when calculating the truncation thresholds. The clean tags obtained were clustered to generate operational taxonomic units (OTUs) at a 97% similarity. The taxonomic identities of the soil microbial communities were checked using the Silva database (Release 138, http://www.arb-silva.de). The diversity and composition of the bacterial and fungal communities under different treatments were determined based on the instructions Caporaso et al. (2010). The Chao1 and ACE (species abundance), as well as the Shannon and Simpson indices (community diversity), were used to estimate the fungal and bacterial alpha diversity (within samples). Krona charts were used to show the abundances of each bacterial and fungal taxa graphically. For Beta diversity, R (Version 3.3.1) was used to visualize the similarity distance using Principal Coordinate Analysis (PCoA). Selected soil physicochemical properties were used to establish the identity and the association of the most abundant bacterial and fungal phylum using partial-RDA.

Soil physicochemical characteristics

All the treatments containing FM, irrespective of its proportion, had significantly higher values (p < 0.05) of the measured soil chemical properties. There was no significant difference in NH₄⁺ concentration among the FM2 and FM3. However, FM1 had a significantly higher NH₄⁺ concentration among all the treatments evaluated. NO₃⁻ had lower values in the FM treatments when compared to the control (CK) (Table 1). While FM increased the soils’ pH relative to CK, there was no significant change in the soil pH among the FM treatments. Also, the highest AK, EC, TN, TON, and TOC values were obtained in FM3. Also, the AK, TOC, and TON values were significantly highest and statistically similar among FM2 and FM3 treatments. Additionally, the available P content was statistically higher in the FM treatments compared to the CK, although it was highest in value in the FM2.

Soil enzyme activity

The influence of FM proportion on the soil enzyme activity: cellulase, phosphatase, urease, and β-Glucosidase is presented in Figs. 1A–1D. Our observations showed that there was no significant (p < 0.05) effect of FM rates on cellulase activity (Fig. 1A). However, the control treatment had a significantly higher cellulase activity than the FM1 and FM2 while being statistically similar to FM3. On the other hand, phosphatase, urease, and β-Glucosidase enzymes were stimulated by FM relative to the control. Among the FM proportions, FM3 gave the highest (p < 0.05) activity of phosphatase and urease (Figs. 1B, 1C). While there was no significant difference in the activity of β-Glucosidase among the FM proportions evaluated, its activity was higher in the FM amendments than the control (Fig. 1D).

Sugarcane agronomic characteristics

The application of FM, irrespective of its proportion, significantly increased the plant height, number of tillers, and plant stem diameter compared to the control treatment (Figs. 2A, 2D and 2E). The use of FM1 gave the significantly highest (p < 0.05) plant height, among other FM treatments. Similarly, FM1, while statistically similar (p < 0.05) to other FM treatments, gave higher values for the number of tillers (Fig. 2D) and plant diameter (Fig. 2E). However, FM treatments had no significant difference (p < 0.05) when compared to the control with respect to the leaves’ chlorophyll content (Fig. 2B) and the number of leaves (Fig. 2C). There was no statistical difference in the leaves’ chlorophyll content and the number of leaves among the FM treatments, despite the observed reduction in FM2 and FM3.

Relative abundance of dominant microbial phyla in filtered mud amended soil

The relative abundances of the top 10 soil bacterial and fungal phyla as influenced by the different FM:soil ratios are presented in Figs. 3A, 3B. The Proteobacteria was dominant in the soil, and their relative abundance ranged from 42.54–50.76% and decreased with an increasing rate of FM. The Actinobacteria (14.74–25.59%), which were the second most abundant bacterial phyla, increased in the FM treated soil relative to the control. Its highest proportion was obtained in FM3. These were followed by the Acidobacteria (4.51–11.48%), which decreased in the FM treated soils. Among the phyla observed, Bacteroidetes (3.65–6.03%), Chloroflexi (2.7–4.71%), Planctomycetes (1.16–3.49%), and Candidatus Saccharibacteria (0.43–2.83%) increased in the FM treated soils compared to the control. However, the relative abundances of the Gemmatimonadetes (3.44–5.9%) and Firmicutes (1.52–2.27%) also reduced in the FM treated soils compared to the control (Fig. 3A). Similarly, there was an alteration in the relative abundances of fungal phyla after the FM application (Fig. 3B). The most dominant fungal phyla observed were the Ascomycota (29.73–59.13%), whose relative abundance was reduced in the FM treated soils compared to the CK treatment. Its lowest abundance was observed in FM2, among the FM-treated soils.

The Alphaproteobacteria class dominated the Proteobacteria class members and was reduced in FM treatments relative to the control. Its lowest proportion was observed in FM3 (Fig. 4A). However, there was a notable increase in the relative abundance of the Betaproteobacteria and Gammaproteobacteria, which comprise members that are ammonium oxidizing due to FM. This was confirmed in the increase in the ammonium oxidizing genera, Devoisa, Luteimonas, and Povalibacter in the FM treatments (Fig. S1). Similarly, bacterial class distribution revealed the abundance of Actinobacteria, which contains N-cycling genera, such as the Nocardioides in FM treatments (Fig. S1), with FM 3 having the highest relative abundance (Fig. 4A). Gemmatimonadetes and Sphingobacteriia reduced in abundance in FM treatments relative to the control. The FM amendment suppressed the abundance of the fungal class Sordariomycetes and Eurotiomycetes (Fig. 4B). While the Sordariomycetes had their highest proportion in the FM2 treatment, there was no significant difference in the relative abundance of Eurotiomycetes among the FM treatments. The fungal class Dothideomycetes was, however, stimulated in FM, especially in FM1 and FM2. The Saccharomycetes were increased under FM, with their highest proportion in FM2.

Alpha and beta diversity of soil bacterial and fungal communities

Across the diversity indices evaluated, the FM amended treatments significantly reduced the diversity (Shannon and Simpson) (Figs. 5C, 5D) and richness matrices (Chao1) (Fig. 5B) based on the observed number of OTUs. The least bacterial diversity across these indices was observed in the FM3 treatment. According to the ACE estimator, the FM3, while having the least diversity, was significantly similar to all other treatments (p < 0.05) (Fig. 5A). No significant difference (p < 0.05) was observed in the fungal community richness indices, ACE (Fig. 5E), Chao1 (Fig. 5F), and Shannon (Fig. 5G) among the treatments. However, Simpson’s diversity index showed that FM2 had the least fungal diversity, while FM1 gave the highest diversity (Fig. 5H).

The beta diversity analyses represented by the principal coordinate analysis (PCoA) (Figs. 6A, 6B) showed the shift of bacterial and fungal community composition under different FM:soil ratios. Beta diversity among the soil samples was determined based on the unweighted Unifrac distance matrices. The PCoA of the bacterial taxa showed that the bacterial communities in FM1, FM2, and FM3 soil samples were clustered and separated from the CK soil sample. This indicates that the FM amendment significantly changed the soil microbial composition (Figs. 6A, 6B).

Similarly, the PCoA of the fungal taxa revealed an indistinct clustering pattern among the treatments (Fig. 6B). It was, however, observed that the control treatment was distinct from FM treatments (Fig. 6B). Distinct clustering of fungal communities in FM2 and FM3 was also observed in the PCoA.

Relationship between soil properties and the distribution of bacterial and fungal communities

Redundancy analysis (RDA) was performed to explore the influence of soil variables on the relative abundance and community composition of bacterial and fungal phyla (Figs. 7A, 7B). It was observed that RDA1 and 2 accounted for 67.07% and 12.32%, respectively, for the changes in the bacterial phyla composition, while RDA1 and 2 accounted for 42.14% and 11.18%, respectively, for the differences in fungal community composition (Fig. 7B). The content of soil AP, pH, TOC, and TON positively influenced the abundance of Bacteroidetes and Chloroflexi, while the availability of NH₄⁺ and TON correlated with the relative abundance of the Actinobacteria. The concentration of soil NO₃⁻ showed a close association with Acidobacteria and Gemmatimonadetes. Similarly, the availability of NO₃⁻ influenced the abundance of Ascomycota (Fig. 7B). Likewise, TN, TON, pH, C/N ratio, and AP were the significant factors that influenced the abundance of Basidiomycota. The TOC of the treatments influenced the abundance of Rozellomycota.