A controlled-release fertilizer did more than feed sugarcane. At one application rate, it was linked to higher sugar yield while reshaping the root-zone microbes and metabolites surrounding the plant. Root–soil underground interactions mediated by soil microorganisms and metabolites are crucial for fertilizer utilization efficiency and crop growth regulation.

The study combined soil microbial community profiling and non-targeted metabolomics to investigate root-associated microbial aggregation and metabolite mechanisms under varying controlled-release fertilizer application rates. The experiment included five controlled-release fertilizer rates, from D1 at 675 kilograms per hectare to D3 at 2025 kilograms per hectare, plus traditional fertilizer as the control.

The optimal field application was D25, producing a 16.3 to 53.6 percent increase in sugarcane yield compared with the control. Sugarcane requires a large and continuous fertilizer supply because of its long growing season. That requirement increases cultivation costs and can reduce economic benefits for farmers.

Controlled-release fertilizer releases nutrients slowly and continuously to support plant growth. Compared with traditional fertilizers, it releases nutrients at a stable rate through special packaging materials, meeting plant growth needs while reducing nutrient wastage and environmental pollution.

The study therefore focuses on CRF’s applicability and mechanism in sugarcane, including interactions between soil metabolites and microorganisms. Soil microorganisms and metabolite–plant interactions influence the acquisition of high-yielding plant phenotypes.

Functional soil microorganisms can promote plant growth through plant growth hormones that regulate root morphology and nutrient uptake. Microbiota also release nitrogen, phosphorus, and potassium by decomposing organic matter and dissolving minerals. Sugarcane produces abundant secondary metabolites, including sugars, ketoacids, and amino acids, which are associated with growth, stress adaptation, and rhizosphere microbial interactions.

The study investigates CRF effects on sugarcane growth using high-throughput soil sequencing and non-target metabolome analysis. Its first question is the optimal CRF application and the composition and diversity of rhizosphere microbial communities during cultivation. The field experiment used sugarcane variety ROC22 and a controlled-release fertilizer with an N:P:K ratio of 18-8-14.

The conventional fertilizer control was uncoated and contained the same nitrogen, phosphorus, and potassium nutrients. Before the experiment, the soil pH was 4.45, and soil organic matter was 43.10 grams per kilogram. A randomized block design compared five film-coated CRF rates with conventional fertilizer as the control.

The CRF rates were 675, 1012.5, 1350, 1687.5, and 2025 kilograms per hectare, labeled D1, D15, D2, D25, and D3. Each experimental treatment was replicated three times, while the control was replicated five times. Sugarcane growth was significantly affected by the CRF application rate, with a p-value below zero point zero five.

Plant height and stem diameter peaked under D25 and D3 conditions, respectively, while sucrose content peaked under moderate D25 conditions. The highest cane stem yield occurred at the highest CRF rate, but sugar yield, the ultimate objective, corresponded to D25.

For superior sugar yield and economic returns, the effective application threshold was found under D25 conditions. Table one compares six CRF application treatments across sugarcane growth, composition, and yield indicators, with values reported as means plus or minus standard deviations.

The letter groupings mark significant differences at p less than zero point zero five: D25 shows the highest listed sucrose content and sugar yield, while D3 has the highest cane yield and stem diameter. This matters because the table reveals that fertilizer rate affects both individual cane traits and population-level yield outcomes.

Different CRF levels significantly affected conventional nutrient content in sugarcane rhizosphere soils. Compared with the control, D1 and D15 significantly increased soil pH and reduced soil acidity. In D15, soil organic carbon peaked, and available nitrogen showed a corresponding trend.

Available phosphorus decreased across the D1 to D2 range, while maximum available potassium corresponded to maximum CRF application. Table 2 shows how controlled-release fertilizer application rates affected sugarcane rhizosphere soil nutrients. The D15 treatment had the highest reported soil organic carbon, at 56.06 ± 9.64 grams per kilogram, and available nitrogen, at 169.17 ± 67.26 milligrams per kilogram, while available phosphorus and potassium varied across treatments, with the highest values under CK and D3, respectively.

The lowercase letters mark statistically significant differences at p less than 0.05, making the table useful for connecting fertilizer rate with changes in soil acidity and nutrient availability. Bacterial ACE diversity differed significantly between D25 and D1, while Chao1 differed among D25, CK, and D1.

Fungal Shannon, ACE, and Chao1 diversity remained largely unchanged despite fertilizer-rate changes. The major bacterial groups included Proteobacteria, Actinobacteriota, and Acidobacteriota, among others. The dominant fungal groups included Glomeromycota, Chytridiomycota, Mortierellomycota, Rozellomycota, Basidiomycota, and Ascomycota.

Rhizosphere metabolites were mainly annotated to secondary-metabolite biosynthesis, carbohydrate metabolism, energy metabolism, and lipid metabolism pathways. Figure one compares rhizosphere bacterial and fungal phylum-level relative abundances across sugarcane CRF treatments, alongside the top twenty KEGG-annotated metabolic pathways in soil metabolites.

The pathway bars report both metabolite counts and their percentage of all annotated metabolites, with biosynthesis of type two polyketide products comprising ninety-three metabolites, or four point eighty-eight percent. Together, these panels show how CRF conditions are evaluated in relation to microbial community composition and metabolic functions.

NMDS showed complete separation of microbial samples under different CRF application conditions. For the metabolic set, principal coordinates analysis showed that principal coordinate one explained 74.71 percent. All treated replicate samples were discriminatory based on metabolite expression.

Under stringent LDA screening, bacterial groups including Gammaproteobacteria, Acidimicrobia, and Bryobacter distinguished biomarkers of CRF application. The fungal groups included Waitea, Ente-rocarpus, Herpotrichiellaceae, Saccharomycetales, and Mycena noctilucens.

The number of significantly altered metabolites varied across CRF application comparisons. D25 versus D15 had twenty-four changed metabolites, while D3 versus D1 had fifty-seven. D2 versus CK had the highest total, with 102 altered metabolic markers; eighty-four increased and eighteen decreased.

Figure two combines LDA bar charts with metabolite radar charts to show which rhizosphere microbial genera and metabolites distinguish fertilizer applications. The bacterial and fungal panels use an LDA threshold above four and p below zero point zero five, while the radar charts label the top ten significantly up- or down-regulated metabolites and plot log-two fold change.

For the D2 versus D3 comparison, the authors highlight metabolites including taurocholic acid and CMP-N-acetylneuraminate among those increased, and alpha-linolenic acid and prostaglandin F2alpha among those decreased, linking CRF application to coordinated microbial and chemical shifts.

Redundancy analysis at the microbial genus level and for metabolites revealed key drivers of the rhizosphere microenvironment. In the bacterial compartment, soil pH, available nitrogen, and soil organic carbon were strongly correlated with bacterial community distribution.

Figure three maps how replicate samples under different controlled-release fertilizer conditions relate to soil nutrients and selected microbial or metabolite features. Panels a and b show bacterial and fungal genus-level redundancy analyses, while panel c presents metabolite RDA; arrow direction indicates correlation and arrow length its magnitude.

The orange labels identify retained top-abundance features, including taxa associated with pH, ammonium nitrogen, and soil organic carbon, such as Proteobacterium, Acidothermus, Trechispora, and Trichoderma. Random forest screening identified the top twenty features important for variability among CRF applications.

Anaeromyxobacter had high abundance under D1 and D25, but lower abundance under D3 and in the control without CRF. Leotiomycetes and Cercospora increased with increasing CRF application, reaching their highest abundance under D3. The characterized metabolites included isoleucyl-proline and methylmalonic acid.

Figure Four uses randomized forest screening to display the top twenty bacterial genera, fungal genera, and metabolites associated with variability among controlled-release fertilizer applications. In each panel, the horizontal position represents mean importance, while the adjacent heatmaps show how abundance varies between treatments.

The authors use this combined view to demonstrate that CRF application affected the sugarcane rhizosphere’s microbial and metabolite composition and to identify features for examining relationships across these groups. Heatmaps were used to illustrate correlations among bacteria, fungi, and metabolites.

The microbial populations of sugarcane rhizosphere soils had significant concomitant relationships. Sphingomonas showed a significant negative correlation with several dominant rhizosphere fungal genera, including Penicillium, Coprinellus, Aspergillus, Cladosporium, Candida, and Mortierella.

Reyranella showed a significant negative regulatory relationship with Cladosporium, Candida, and Mortierella, and a significant positive relationship with Mycena. Burkholderia, Bradyrhizobium, and Bryobacter showed a significant negative correlation with Xanthoxic acid, while Plectosphaerella showed a significant positive correlation with it.

These numerous and complex relationships do not indicate causality, but they can provide a foundation for exploring molecular regulatory mechanisms. Figure five maps Spearman correlations across the bacterial–fungal–metabolite triad: bacteria–fungus in panel a, bacteria–metabolites in panel b, and fungi–metabolites in panel c.

Green represents positive associations and blue negative ones, with marked cells indicating statistically significant relationships; the analysis focuses on the top thirty microbial genera and top twenty metabolites by correlation degree. Notably, Sphingomonas shows negative correlations with several dominant fungi, while Mycobacterium, Conexibacter, and Acidipila Silvibacterium show positive correlations with more fungi, highlighting potential microbial interactions linked to rhizosphere homeostasis.

The application rate of controlled-release fertilizer had a significant impact on sugarcane growth in the field. Different CRF rates produced varying effects on plant height, stem diameter, and sucrose content, creating trade-offs between cane stem yield and sugar yield.

To optimize sugar yield and economic returns, the recommended condition is D25. ACE and Chao1 indices revealed differentiation between treatments, suggesting that CRF application can modulate bacterial richness. Fungal alpha diversity indices remained stable across fertilizer rates.

The rhizosphere bacterial communities were dominated by Proteobacteria, Actinobacteriota, Firmicutes, and other key groups. CRF significantly affected the composition of microorganisms and metabolites in sugarcane rhizosphere microcosms, as shown by random forest analysis.

Anaeromyxobacter was abundant under D1 and D25, decreased under D3 and in the control, and may have a potential role in nutrient cycling and adaptation to different CRF levels. Leotiomycetes and Cercospora increased with CRF application and reached their highest levels under D3.

Isoleucyl-proline, seven-deoxyloganetate, lolitrem E, nicotinamide, and methylmalonic acid showed higher abundance with CRF than in the control without CRF. The study identifies D25 as the most effective controlled-release fertilizer condition for sugar yield and connects that outcome with changes in soil nutrients, microbial communities, and metabolites, while emphasizing that correlations do not prove causality.