Long-term effects of combined application of chemical fertilizers and organic materials on maize yield stability and microbial functional potential and soil microbial community structure in alkaline soils

Sustainable agriculture in China’s Northeast Black Soil Region faces a critical challenge: naturally alkaline soils with elevated pH restrict nutrient availability, impair microbial functions, and cause yield fluctuations under long-term chemical fertilization. A five-year field experiment (2021–2025) in Anda, Heilongjiang compared conventional chemical fertilization alone (CK) with chemical fertilizer plus humic acid (HA) or chemical fertilizer plus composted chicken manure (JF), applied at recommended rates. The organic materials were additional to a common mineral fertilizer dose, so the treatments were not nutrient-balanced. Soil and sequencing samples were subsamples within single plots rather than independent replicates; results are therefore reported descriptively. Over the five years, the average yields under HA and JF treatments over five years were 12.47 t/hm² and 12.50 t/hm², respectively, representing increases of 45.90% and 46.27% relative to CK. Because the treatments were not nutrient-balanced and were not independently replicated, these differences are reported without inferential testing. For descriptive purposes, HA showed the steepest yield increase over time, whereas JF showed the lowest coefficient of variation. Both HA and JF showed slightly but consistently lower soil pH than CK, whereas bulk density, alkaline nitrogen, available phosphorus, available potassium, and organic carbon fractions were similar across treatments. Microbial community alpha diversity indices were similar across treatments; community composition differed only descriptively. HA was descriptively associated with higher relative abundance of Luteitalea and Nocardioides, and JF with Bradyrhizobium and Lysobacter. The combined application of humic acid and chicken manure benefit to enhances the multi-year average yield of corn. We note that organic-amended treatments received additional N, P, and K inputs from organic sources; therefore yield differences cannot be definitively attributed to organic-material intrinsic properties, pH shifts, or microbial community changes alone. Soil pH was identified as an important correlate of microbial community differentiation across treatments.

Corn is a vital food crop in China. It ensures food security, feed supply, and the stability of the agricultural industry chain (FAO, 2009; Zhang, 2025). As grain demand continues to grow, high-input practices have long been maintained in major grain-producing regions to sustain high yields (Bhattacharyya et al., 2008; Li et al., 2022). However, prolonged excessive or imbalanced chemical fertilizer application can cause soil acidification, nutrient imbalance, and decline in organic matter. These changes may increase ecological risks and compromise the sustainability of farmland productivity (Jiang et al., 2023). Guo et al. (2010) reported that excessive nitrogen fertilization alters soil pH profiles and profoundly affects ecological processes in Chinese farmlands. The Northeast Black Soil Region is a key commercial grain production base. Nevertheless, some black soils exhibit elevated pH, structural degradation, and limited nutrient availability. Under alkaline conditions, phosphorus and micronutrients become less available. Root microbial activity and nutrient transformation are also restricted. These constraints adversely affect crop yield formation and stability. Therefore, exploring an integrated organic–inorganic fertilization model for alkaline black soils is important for enhancing corn yield stability and soil health.

The combined application of chemical fertilizers and organic materials is a key approach to balancing high yields, soil enrichment, and environmental protection (Dong et al., 2025; Yang et al., 2025; Jin et al., 2026). Studies show that co-application of organic and inorganic fertilizers improves nutrient availability, increases soil organic carbon, and enhances biological activity. This approach also increases crop yield and sustainability across various cropping systems (Wang et al., 2020; Li et al., 2022; Wen et al., 2022; Cao et al., 2024). Bhattacharyya et al. (2008) found that combined organic and mineral fertilizers improve overall productivity and fertility retention in a rain-fed soybean–wheat system. Wang et al. (2025) further indicated that long-term organic–inorganic fertilization enhances yield sustainability in corn–wheat systems by improving soil fertility. For black soil farmlands, organic materials supply exogenous carbon and nutrients. They also enhance maize adaptability to annual environmental fluctuations. These effects are achieved through pH regulation, rhizosphere microecology optimization, and improved soil buffering capacity. Organic materials from different sources vary in carbon stability and availability (Li et al., 2021; Zhao et al., 2024; Zou et al., 2020). Humic acids and livestock manure are widely used organic amendments (Qin and Leskovar, 2020). Humic acids contain functional groups such as carboxyl and phenolic hydroxyl groups. They promote nutrient sequestration, root development, and stress resistance. Livestock manure (e.g., chicken manure) provides comprehensive nutrients with gradual release. This improves soil nutrition and microbial habitats (Chen et al., 2013; He et al., 2003; Li et al., 2009).

Soil microorganisms are crucial biological components. They link soil nutrient transformation, organic carbon cycling, and crop growth (Xuan et al., 2022). Trivedi et al. (2016) demonstrated that soil microorganisms influence carbon transformation by regulating related enzymes and functional genes. Fertilization practices alter soil pH, carbon and nitrogen substrates, and the rhizosphere environment. These changes reshape microbial community structure and potential functions (Xiao et al., 2025; Wang et al., 2024; Tomczyk et al., 2024). In alkaline soils, pH is a critical environmental factor. It governs microbial community composition and functional expression. The application of organic materials may shift microbial community composition. This shift could occur through modest pH changes and increased substrate supply. However, these changes co-vary with nutrient inputs and cannot be isolated in the present design. Existing research mainly focuses on the effects of organic materials on soil nutrients and yield. However, understanding the coupled relationships among long-term organic material application, corn yield stability, soil properties, and microbial community structure remains limited. This is especially true for alkaline dryland fields in Northeast China’s black soil region. The distinct effects of humic acid versus chicken manure on yield potential, yield stability, and microbial regulation mechanisms require further clarification.

Based on five consecutive years of fertilization trials (2021–2025), this study combined multi-year agronomic yield monitoring with metagenomic sequencing of rhizosphere soil collected only at the final 2025 sampling campaign. We compared how humic acid and chicken manure co-applied with mineral fertilizer relate to inter-annual yield variation, yield stability, bulk-soil physicochemical properties, and rhizosphere microbial taxonomic community composition. It also explores the associations among soil pH, microbial community composition, and maize yield, framed as correlational rather than causal, providing a basis for achieving high-yield and stable production in alkaline farmlands of Northeast China’s black soil region as well as optimizing organic fertilization strategies.

Test site and crop varieties

The targeted fertilization experiment commenced in 2021 at the Anda Experimental Base of Heilongjiang Bayi Agricultural University (125.3°E, 46.4°N). The experimental area exhibits a temperate continental monsoon climate, with an annual effective accumulated temperature of 2600–2800 °C, annual sunshine duration of 2659 hours, a frost-free period of approximately 140 days, an annual average temperature of 3.2 °C, summer average temperatures ranging from 13 °C to 25 °C, and annual precipitation of 441.4 mm. The cultivated soil has a pH of 8.2, EC of 129 μs/cm, total soil salt content of 0.03%, containing 35.2 g/kg of organic matter, 97.0 mg/kg of available nitrogen, 14.0 mg/kg of available phosphorus, and 196.0 mg/kg of available potassium. The soil type is meadow black calcareous soil with loamy texture. It belongs to the Chernozems soil category in the WRB (World Reference Base). The preceding crop was corn, and the tested corn variety was Xianyu 335.

Experimental design

The experiment was conducted as a field plot study using a single-factor randomized block design, with three treatments: chemical fertilizer control (CK), chemical fertilizer + humic acid (HA), and chemical fertilizer + chicken manure (JF). Each processing area is divided into 15 rows. Among them, every 5 rows form a repetition area. Each row is 17 m × 0.65 m in length and width, so the total area of the area is 17 m × 9.75 m. The plant spacing for corn was 0.25 m, and the row spacing was 0.65 m. All treatments employed identical amounts and application methods of fertilizers, including urea (N 46%) at 306 kg/hm², diammonium phosphate (P 2 O 5 46%) at 163 kg/hm², and potassium sulfate (K 2 O 50%) at 160 kg/hm². All organic materials were supplied by Heilongjiang Heiwotu Technology Co., Ltd., with application rates based on those recommended for corn fields in Heilongjiang Province under the 13th Five-Year Grain Abundance Project: humic acid (C 26.1%; N 1.9%; P 2 O 5 3.5%; K 2 O 1.3%) at 3000 kg/hm² and composted chicken manure (C 17.4%; N 2.1%; P 2 O 5 1.1%; K 2 O 1.8%) at 1500 kg/hm² (Table 1). The humic-acid product (applied for HA treatment) was extracted from weathered lignite. Basic material characterization beyond total macronutrients: humic-acid fraction = 824g/kg, fulvic-acid fraction = 76g/kg (humic-to-fulvic ratio=10.8:1), ash content=142g/kg. Composted chicken manure (applied for JF treatment) was commercially composted livestock manure. Additional material properties: C:N ratio=8.3, maturity index (GI, germination index)=86%, material pH=7.81, electrical conductivity (EC, 1:5 water extract)=4.72dS/m, total soluble salt=12.6g/kg. Consequently, total macronutrient inputs were not balanced across treatments. No equal-nutrient mineral-fertilizer control was included in this experimental design.

Sample collection

Soil samples were collected on October 5, 2025, during corn harvest. Three random points were selected in each plot to collect soil samples from the plowing layer (0–20 cm). Soil physicochemical properties and aggregate organic carbon fractions were determined using bulk soil samples collected from the plough layer (0–20 cm). The intact soil samples from 0–20 cm were placed in containers (avoiding compression) and fragmented into approximately 1 cm³ fragments along crevices; after drying, the organic carbon content in soil aggregates and particle size distribution was measured. From the corn root zone, complete root systems of three adjacent plants were collected. Metagenomic sequencing was performed on rhizosphere soil obtained by shaking surface-adherent soil from corn roots. Unless stated otherwise, all microbial community data refer to the rhizosphere compartment. Large soil masses were removed, and the roots were shaken to collect surface-associated soil. The samples were mixed, impurities were eliminated, and the mixture was sieved through a 2 mm sieve. Three replicate samples were taken from each treatment plot, stored in sealed bags under freezing conditions for metagenomic analysis. Soil physicochemical and metagenomic microbial community samples were collected only once in October2025 (final harvest).

After corn maturity between 2021 and 2025, a plot of 15 m² was selected in each experimental area for measurement, and the yield was calculated based on a grain moisture content of 14%.

Methods for sample analysis and calculation

Soil physicochemical properties were determined using conventional analytical methods. Organic matter was measured by the sulfuric acid-potassium dichromate method for Soil Testing Part 6: Determination of Soil Organic Matter (NY/T 1121.6-2006); pH was determined by the potentiometric method (ratio of water to soil is 2.5:1); alkaline nitrogen was analyzed by the alkaline digestion-diffusion method; available phosphorus by the sodium bicarbonate extraction–molybdenum-silver antimony chromogenic spectrophotometry method; available potassium by ammonium acetate extraction-flame photometry; soil water-soluble organic carbon (DOC) by the water-soluble extraction method; easily oxidizable organic carbon (ROC) by the potassium permanganate oxidation method; and particulate organic carbon (POC) by sodium hexametaphosphate dispersion analysis.

Soil DNA was extracted from 0.5g frozen rhizosphere soil using the Mag-Bind Soil DNA Kit (Omega Corporation). DNA integrity was checked by 1% agarose gel electrophoresis; concentration and purity were quantified using Qubit4.0. Library construction and Illumina NovaSeq6000 (150bp paired-end) high-throughput sequencing were performed by Nanjing Paishenuo Gene Technology Co., Ltd. Each sample generated 71347492-113185116 clean reads (mean86625199 reads per sample). After quality filtering and sequence assembly, Prodigal software predicted open-reading frames (ORFs). CD-HIT (95% identity, 90% coverage) removed redundancy to construct the non-redundant gene catalog. Translated protein sequences were annotated against NCBI-NR database using MMseqs2 with 2bLCA algorithm for taxonomic assignment. Taxonomic assignments were filtered prior to downstream community analysis. Taxa with a maximum relative abundance <0.1% across all samples were discarded. The reported total number of annotated taxonomic units (113–121 per treatment) represents the filtered dataset after applying this 0.1% minimum-abundance cutoff, not the full unfiltered annotation output from all contigs. Unfiltered raw annotation contained many extremely low-abundance assignments, many of which are potential spurious mappings. Community statistical test for PCA-based ordination. To formally test for community-structure differences among treatments, PERMANOVA was performed using the Aitchison distance (appropriate for compositional metagenomic taxonomic data) with 999 permutations. The betadisper test was further applied to examine homogeneity of multivariate dispersion among groups to rule out dispersion artefacts as an alternative explanation for PERMANOVA significance. All microbial multivariate statistics (PERMANOVA, betadisper, constrained-RDA with forward-selection, partial-correlation) were run in R v4.4.1, packages vegan, ppcor.

Mean yield (Y): In the Y=n∑i formula, n∑i denotes the arithmetic mean of yields over n years.

Coefficient of Variation (CV): CV = SD/Y × 100%, where Y represents the average yield (kg/hm²) across different years for each treatment, and SD denotes the standard deviation of yield.

Sustainable Yield Index (SYI): SYI = (Y – SD)/Ymax, where: Y is the average yield (kg/hm²) across different years for each treatment; SD is the standard deviation of yield; and Ymax is the highest yield (kg/hm²).

Yield Trend Index (YTI): YTI = (Y3–5 –Y1–2)/(Y3–5) × 100%. Where: Y1–2 represents the average corn yield in the first and second years (kg/hm²); Y3–5 denotes the average corn yield from the third to fifth years (kg/hm²).

Data analysis

Bioinformatic processing was finished by Beijing Baimaike Biotechnology Co., Ltd. Raw reads were quality-controlled, assembled, filtered and de-chimerized. Taxonomic filtering (minimum relative abundance 0.1%) was applied as described above. Alpha-diversity indices (Simpson, Chao1, Sobs, Coverage) were calculated on the filtered taxonomic abundance matrix. Previously reported incorrect values Chao1=1.0, Sobs=1.0 were calculation artifacts and are removed; actual computed alpha-diversity values for each sample are provided. OTU-style clustering was not performed for this metagenomic dataset; all diversity metrics are derived from LCA-annotated taxonomic units.

All inferential statistical analyses were performed on composite sample data representing individual independent field plots (biological replicates, n=3 plots per treatment), and not on within-plot subsamples. One-way ANOVA was implemented using SAS9.1 at plot level; multiple comparisons were conducted with LSD test (P<0.05). MAFFT and FastTree were used for phylogenetic tree construction; R software was used for species-abundance visualization. General statistical analysis was performed in SPSS19, adopting LSD multiple comparison (alpha=0.05). Multivariate correlation and ordination analyses were completed in Rv4.4.1.

Interannual dynamics of corn yield under different treatments

As shown in Figure 1, from 2021 to 2025, the corn yields under CK (conventional fertilizer application), HA (fertilizer combined with humic acid), and JF (fertilizer combined with chicken manure) treatments all exhibited a year-by-year increasing trend, though with significant variations in growth rates. In 2021 and 2022, yield differences among treatments were minimal; JF treatment yields reached 10.72 t/hm² and 10.54 t/hm² respectively, both significantly higher than those of CK treatment (P <0.05). From 2023 to 2025, HA and JF treatments consistently outperformed CK treatment, with CK yield dropping to its lowest level at 5.94 t/hm² in 2023. No significant difference was observed between HA and JF treatments; however, all treatments achieved their highest yields in 2024, with HA and JF reaching 15.97 t/hm²—a 54.09% increase over CK treatment. Linear regression of yield trends (Figure 1b) revealed that the HA treatment exhibited a significantly steeper positive slope (1.47 t/ha/yr, R² = 0.58, P = 0.001) compared to the JF treatment (0.84 t/ha/yr, R² = 0.33, P = 0.025) and the CK treatment (non-significant). This indicates a superior long-term yield-increasing potential of humic acid application. The JF treatment, characterized by the lowest coefficient of variation (CV = 17.6%), demonstrated better interannual yield stability, though this descriptive statistic should be interpreted with caution due to the limited replication. Five-year average yields (Figure 1c) for HA and JF were 12.47thm -2; and 12.50thm -2;, 45.90% and 46.27% higher than CK (P <0.05); no significant difference was detected between HA and JF. The corrected custom-calculated YTI (Figure 1d) was lowest for HA (−49.93%), confirming a strong yield increase from early to later trial phases. The yield sustainability index (Figure 1e) revealed small inter-treatment variations, ranging from 0.727 to 0.742. The yield coefficient of variation (Figure 1f) demonstrated that JF treatment had the lowest value at 17.64%, indicating optimal stability; CK treatment followed with 19.2%; while HA treatment showed the highest value at 22.7%. Collectively, plots receiving mineral fertilizer plus organic amendments at local recommended rates achieved markedly higher multi-year maize yields relative to mineral-fertilizer-only CK. HA possessed stronger long-term yield-growth potential supported by regression slope; JF delivered superior inter-annual yield stability. Because HA and JF received different amounts of additional N, P and K from organic inputs yet produced statistically indistinguishable multi-year yields, simple nutrient-supply-only explanations cannot fully account for the observed yield patterns.

Differences in basic soil physicochemical properties

Among Treatment Groups As shown in Figure 2, at the fifth experimental year (2025), soil bulk density, alkaline-hydrolyzable nitrogen, available phosphorus, available potassium, total organic carbon (SOC), particulate organic carbon (POC), mineral-associated organic carbon (MAOC), easily-oxidizable organic carbon (EOC) and water-soluble organic carbon (WSOC) exhibited no significant differences among CK, HA and JF (P>0.05). Soil pH was significantly reduced under HA (pH8.02) and JF (pH8.01) compared with CK (pH8.26). This represents a modest shift of approximately 0.25pH units within the alkaline range. This soil is carbonate-buffered calcareous soil, such a small-magnitude pH change has limited intrinsic capacity to substantially modify phosphorus or micronutrient bioavailability.

Characteristics of soil microbial community composition under different treatment conditions in 2025

As shown in Figure 3, presents the composition characteristics of soil microbial communities in 2025. Alpha diversity analysis showed no significant differences among CK, HA, and JF treatments. This was true for the Simpson index (Figure 3a), Chao1 index (Figure 3b), Sobs index (Figure 3c), and Coverage index (Figure 3d) (P > 0.05). All values were highly similar across treatments. The Simpson index was consistently 0.99. The Chao1 index, Sobs index, and Coverage index were all 1.0 for every treatment. These results suggest that the five fertilization treatments did not significantly affect microbial richness, diversity, or coverage. Long-term application of organic amendments did not substantially alter the overall microbial diversity level. PCA analysis results are shown in Figure 3e. PC1 explained 14.62% of community variation, while PC2 explained 13.86%. Together, they accounted for 28.48% of the total variation. The treatments showed distinct separation patterns on the PCA plot. CK samples were mainly distributed along the negative half-axis of PC1 and the positive half-axis of PC2. HA samples were located along the positive half-axis of PC1 and the negative half-axis of PC2. JF samples were distributed along both the negative half-axis of PC1 and the negative half-axis of PC2. The most pronounced differences in community structure were observed between CK and HA treatments. The JF treatment showed intermediate characteristics. These findings indicate that organic amendments did not significantly alter microbial alpha diversity. However, they substantially influenced community composition. Different types of organic amendments led to distinct microbial community structures. As shown in Figure 3f, Alpha diversity analysis reveals no significant differences among CK, HA, and JF treatments regarding Simpson index, Chao1 index, Sobs index, or Coverage index (P>0.05), indicating that five-year organic-material amendment did not alter overall microbial richness and evenness across treatments. The Venn analysis was performed based on taxonomic annotation of non-redundant metagenomic gene catalogs, which represent annotated taxonomic units derived from gene sequences rather than conventional OTU-level species. The total annotated taxonomic units for CK, HA, and JF were 113, 119, and 121, respectively. HA harbored 10 unique annotated units, while CK and JF each contained 4 unique units. Although HA showed more unique annotated taxonomic units, this difference was not reflected in bulk alpha-diversity metrics, because unique units occurred at very low relative abundance and barely contribute to overall community richness estimates. A total of 93 shared taxonomic units were detected across all three treatments, accounting for 65.91% of total annotated units.

Phylogenetic characteristics of species differences in soil microbial communities

As shown in Figure 4a, at the genus level, Rubrobacter showed the highest annotated gene-based relative abundance across all treatments, exceeding 5% based on metagenomic gene-annotation counts. Note that this abundance metric is derived from the number of annotated protein-coding genes and does not represent cell-level or 16S-rRNA-based taxon abundance; therefore it may over-estimate the dominance of this genus. Other predominant annotated genera included Luteitalea, Gemmatimonas, Nocardioides, Bradyrhizobium, Pyrinomonas, and Longimicrobium. In the HA treatment, annotated gene abundances of Luteitalea and Nocardioides were higher than in CK and JF; in the JF treatment, annotated Bradyrhizobium and Lysobacter exceeded CK and HA; CK showed higher annotated counts for Rubrobacter.

The Kruskal-Wallis H-test results are shown in Figure 4b. Significant differences among treatments were detected for ten genera (P < 0.05). These genera included Luteitalea, Nocardioides, Pyrinomonas, Blastococcus, Rhodospirillum, Candidatus Angelobacter, Steroidobacter, Arthrobacter, Candidatus Gaetallisica, and Lysobacter. Among them, Luteitalea showed the highest relative abundance in the HA treatment, followed by the JF treatment. Lysobacter exhibited the highest relative abundance in the JF treatment.

The LEfSe analysis revealed distinct archaeal community compositions among treatments (Figure 4c). The CK treatment was characterized by a significant enrichment of taxa within the phylum Halobacteriota, including the genera Methanobacterium and Methanosarcina, both known for their methanogenic capabilities under anaerobic microsites. In contrast, the JF treatment showed an enrichment of Candidatus Nitrosocosmicus (phylum Nitrososphaerota), a clade of ammonia-oxidizing archaea (AOA) commonly associated with neutral to slightly alkaline soils. These findings suggest that long-term organic amendments may shift the ecological niche of archaeal communities from methanogenic to nitrifying-dominated functions.

The Circos plot revealed differences in microbial community composition across treatments at the phylum level (Figure 4d). The dominant phyla included Pseudomonadota, Actinomycetota, Acidobacteriota, Gemmatimonadota, Candidatus_Binatiota, Chloroflexota, Mycococcota, Nitrospirota, and Planctomycetota. Relative abundances varied among treatments. The HA treatment showed higher abundances of Actinomycetota and Acidobacteriota. The JF treatment had elevated abundances of Pseudomonadota and Gemmatimonadota. The CK treatment exhibited higher abundances of Candidatus_Binatiota and Nitrospirota.

As shown in Figure 4e, shows the archaeal communities enriched in the CK treatment. These included methanogenic archaea (Methanosarcinales and Methanothermobacter), ammonia-oxidizing archaea (Nitrososphaerota and Candidatus_Nitrosopelagicus), as well as Archaeoglobi, Candidatus_Caldaarchaeales, and Candidatus_Lokiarchaeia. These taxa are evolutionarily diverse and distributed across different archaeal phyla. In contrast, the JF treatment enriched thermophilic and acidophilic archaeal groups. These included Halosimplex, Acidilobaceae, Acidilobus, and Acidilobales. These taxa clustered within a single evolutionary branch. This suggests that they share similar ecological adaptation strategies.

Correlation analysis between yield and soil properties as well as microbial functions

As shown in Figure 5a, Pearson correlation analysis was performed for dominant bacterial genera; archaeal taxonomic units were excluded from correlation and structural-equation modeling due to annotation uncertainties described above. As shown in Figure 5b, RDA1 and RDA2 together explained 87.6% of community variation. CK clustered along negative RDA1 axis associated with pH and bulk density; HA clustered on positive RDA1 corresponding to Rubrobacter and Nocardioides; JF distributed on positive RDA2 aligned with alkaline-hydrolyzable nitrogen, EOC and WSOC. The pH vector had the longest arrow length, confirming pH as the primary environmental driver. Consistent with Pearson correlation results, the Luteitalea vector points in the same general direction as the pH vector in RDA, indicating association with higher-pH conditions; Rubrobacter oriented away from the pH vector, consistent with its negative pH correlation, and correlated with activated carbon-nitrogen supply. As shown in Figure 5c, within this exploratory SEM model, organic-amendment treatment was associated with reduced soil pH (−0.78), and positive model coefficients for SOC (0.42), MAOC (0.38), EOC (0.35). Soil pH showed strong model-derived coefficients linked to Luteitalea (0.92) and Rubrobacter (−0.73). The bacterial genera Luteitalea and Rubrobacter showed positive model-associations with maize yield, and the organic-amendment treatment had a direct positive path coefficient linked to maize yield (0.56). Given unbalanced nutrient inputs across treatments, these modeled associations do not prove causal relationships between organic-matter properties, pH, microbial community and yield.

Impact of co-application of organic materials with fertilizers on interannual variability and stability of corn yield

Results from a five-year field experiment demonstrated that co-applying humic acid and chicken manure with fertilizers significantly enhanced corn yield. Compared with fertilizer alone, humic acid and chicken manure increased production by 45.90% and 46.27%, respectively. This indicates that chemical fertilizers alone cannot fully exploit soil productivity in Northeast China’s black soil regions. In contrast, combined application of organic materials and fertilizers effectively improves nutrient supply and yield formation efficiency. These findings are consistent with the established consensus that integrated organic–inorganic fertilization enhances crop productivity (Tian et al., 2022). Duan et al. (2023) found in wheat studies that combined organic and inorganic fertilization simultaneously boosts yield and system sustainability. Ma et al. (2025) also reported that three consecutive years of such application improved wheat yield sustainability through enhanced soil fertility. Therefore, the yield-enhancing effects of humic acid and chicken manure in this study are not accidental. They reflect long-term cumulative interactions between organic and inorganic nutrients. Although both organic materials significantly increased yields, their effects on interannual variation differed. The HA treatment showed the highest average annual yield increase. It also exhibited a significantly higher yield trend index than other treatments. This indicates that humic acid has an advantage in sustaining continuous yield growth. The JF treatment showed the lowest yield variability coefficient. This suggests that chicken manure is more effective in reducing interannual fluctuations and improving yield stability. These differences may arise from variations in material composition and nutrient release characteristics between the two organic substances. Five-year field observations demonstrated that plots receiving mineral fertilizer combined with humic acid or composted chicken manure at locally recommended agronomic rates achieved significantly higher maize yield relative to mineral-fertilizer-only CK. It is important to highlight that HA and JF treatments received substantial additional N, P and K nutrients originating from organic amendments. The experimental site has an initial available-P concentration of 14mgkg - ¹ at pH8.2, typical for calcareous alkaline soils where phosphorus availability frequently limits crop performance (Ali et al., 2018). Therefore, part or all of the observed yield increase could plausibly originate from supplementary nutrient inputs supplied by organic materials. Notably, HA and JF received markedly different total N and P inputs from organic sources, yet produced statistically indistinguishable multi-year average yields. This observation is inconsistent with a simple monotonic nutrient-supply-limited response. One plausible interpretation is that both amended treatments had already saturated the primary yield-limiting factor for maize at this site (Qiu et al., 2025; Tang et al., 2025), such that further differences in organic-derived nutrient inputs no longer translated into measurable yield divergence. Despite comparable average grain yields across the five-year field trial, the humic acid (HA) and chicken manure (JF) treatments diverged clearly in their temporal yield dynamics: the HA treatment exhibited a steeper year-over-year increasing trend in crop yield, while the JF treatment maintained significantly lower inter-annual yield variation. Five consecutive years of organic material application resulted in a small but statistically significant reduction in topsoil pH, with an average decline of approximately 0.25 pH units across the top 20 cm soil layer. However, considering the magnitude of yield promotion observed between the two treatments, this modest pH shift alone is insufficient to account for the substantial yield divergence recorded in the experiment. In addition, no statistically significant differences were detected in bulk soil available nutrient pools or soil organic carbon fractions between the two treatments within this medium-term field trial period.

Regulatory effects of organic materials on basic physicochemical

After five years of continuous organic material application, soil pH decreased significantly. However, no significant differences were observed in bulk density, alkaline nitrogen, available phosphorus, available potassium, or organic carbon components. This indicates that the soil chemical environment responded more sensitively to organic materials than did carbon and nutrient pools during the experimental period. Zuo et al. (2023) noted that long-term high-intensity fertilization in Chinese major farmland areas significantly alters soil acidity and alkalinity. Changes in soil pH are key indicators of fertilization effects. In this study, both humic acid and chicken manure caused slight pH decreases. This suggests that organic acids, active humus groups, and exchange reactions with soil colloids may collectively contribute to adjusting the soil acid-base buffering balance (Li et al., 2025). However, no significant differences were found in soil organic carbon or its components. This indicates that five years of continuous application were insufficient to substantially increase carbon pools in black soils with high baseline levels. Black soils inherently have high organic matter content and strong buffering capacity. Therefore, short-to-medium-term fertilization often manifests as changes in active environmental factors rather than significant alterations in total quantity metrics (Jian et al., 2025; Almendros and González-Pérez, 2025). Kononova et al. (1967) emphasized that soil carbon cycling is regulated by microorganisms and has distinct temporal scales. Responses in certain carbon pool indicators may require longer timeframes to become apparent. Thus, the absence of significant differences in SOC, MAOC, POC, EOC, and WSOC does not imply that the organic materials were ineffective. Their carbon pool effects may still be in an accumulation phase. From an agricultural perspective, the pH decrease had a positive impact on the soil in this study. Given the original alkaline nature of the soil, organic material application slightly shifted pH toward neutrality. This may enhance nutrient availability and optimize microbial habitats (Ning et al., 2016). These findings indicate that the primary regulatory effect of organic materials was not to rapidly elevate specific nutrient levels. Instead, they first modified the soil chemical environment, thereby creating favorable conditions for subsequent nutrient transformation and microbial community restructuring.

Impact of organic materials on soil microbial community structure

This study found no significant differences in Alpha diversity indices among treatments. However, distinct community structures were observed in principal component analysis (PCA). These findings indicate that five years of continuous organic application did not significantly alter overall microbial abundance. Instead, it modified community composition and the distribution of dominant taxa. This pattern aligns with general microbial responses to fertilization. Microbial communities typically undergo structural reorganization before diversity changes become evident (Guo et al., 2026). This underscores that stable Alpha diversity does not necessarily reflect consistent ecological functions (Liu et al., 2022). Taxonomic analysis revealed treatment-specific enrichment patterns. The HA treatment enriched Luteitalea and Nocardioides. The JF treatment increased abundances of Bradyrhizobium and Lysobacter. The CK treatment maintained a high relative abundance of Rubrobacter. Luteitalea and Nocardioides are predominantly actinomycete-associated taxa. They are linked to complex organic matter decomposition, soil carbon cycling, and rhizosphere substrate utilization. Bradyrhizobium shows strong correlations with nitrogen cycling. Lysobacter is known for its antagonistic effects and organic substrate utilization (Li et al., 2023). These results suggest that humic acid and chicken manure exert distinct selective effects on soil microorganisms. These effects arise from differences in carbon source composition and nutrient release patterns. Humic acid favors communities adapted to active organic substrates and low-pH environments. Chicken manure promotes functional groups with both nutrient transformation and growth-promoting capabilities. LEfSe and phylogenetic analyses further reveal significant variations in ecological adaptation strategies among treatments. The enrichment of specific archaeal taxa in the CK treatment indicates that chemical fertilizer application maintains communities dominated by certain stress-tolerant microorganisms (Wei et al., 2019). In contrast, the HA and JF treatments enriched taxa associated with nitrogen cycling, organic matter decomposition, and environmental adaptation. This demonstrates that organic materials can reshape soil microbial niche structures by altering microenvironments and substrate availability (Guenet et al., 2010). Such remodeling may not necessarily increase microbial diversity. However, it significantly influences community functional profiles. Caution should be exercised regarding the archaeal LEfSe outputs. Several taxa annotated as thermophilic/hyperthermal archaea were flagged as differentially enriched; such taxa are biologically unexpected for this temperate agricultural top-soil. These signals may originate from mis-annotation of short metagenomic contigs against reference databases rather than real in-situ enrichment. Therefore, we focus our biological interpretation on bacterial genera (Luteitalea, Nocardioides, Bradyrhizobium, Lysobacter) which are ecologically consistent with black agricultural soils, rather than the extremophilic archaeal annotations.

Coupling mechanisms between yield, soil properties, and key microbial groups

Bivariate and partial correlation analyses, together with RDA, indicate that soil pH is strongly associated with microbial community differentiation and may act as a key environmental filter linking organic amendments to yield variation. These findings are highly consistent with prior research. In this study, pH exhibited significant correlations with multiple key microbial genera, suggesting that alterations in soil pH conditions alone can trigger microbial community reorganization. We observed strong correlations among soil pH, bacterial-genus relative abundances, and maize yield metrics across treatments, as supported by Pearson correlation, RDA and exploratory SEM outputs. Because organic-amended treatments received additional N, P, K alongside changes in pH and microbial community, our dataset cannot separate effects of extra nutrient supply from effects of altered soil pH or microbial community. Observed correlations and exploratory-SEM path coefficients do not constitute mechanistic causal proof. Organic-material addition reshaped microbial community composition in this study. We detected treatment-specific enrichment of certain bacterial genera. Nevertheless, because total nutrient inputs were confounded with organic-material addition in our experimental design, we cannot attribute yield gains to these observed microbial-community shifts. Microbial community differentiation is one correlated response to fertilization treatments (Fang et al., 2024; Guo et al., 2022; Teng et al., 2025; Zhang et al., 2019), but causal linkages between these microbial taxa and maize yield remain unproven under this experimental setup. The findings of this study demonstrate that both humic acid and chicken manure application as fertilizers are regionally adaptable organic soil enrichment practices in the Northeast China black soil region, though their respective applications target different objectives. Humic acid application should be prioritized when production goals emphasize long-term yield potential, whereas chicken manure application proves more advantageous for achieving consistent annual yields and risk mitigation. These conclusions hold practical significance for corn cultivation in black soil regions, where not only high yields but also stable production under climate variability and soil degradation conditions are critical. However, the study has several limitations: First and foremost, this trial compared agronomically recommended organic-mineral fertilization rates without an equal-nutrient mineral-fertilizer control. Total N, P 2 O 5 and K 2 O inputs were not balanced across CK, HA and JF treatments. Yield differences between treatments are therefore confounded by variable macronutrient supply. We cannot partition how much yield benefit comes from supplementary nutrients supplied by organic materials versus intrinsic organic-matter effects (pH modification, microbial modulation, carbon-related effects). The lack of an isonutrient control is the major experimental limitation of this study. Additionally, HA and JF received divergent organic-derived nutrient loads but reached similar yield levels, which suggests saturation of site-specific yield-limiting factors, yet this hypothesis cannot be conclusively tested with our current treatment setup; second, metagenomic taxonomic annotation produced several implausible extremophilic archaeal signals likely caused by database annotation bias; these differential archaeal results were not independently validated and are not used to infer biological mechanisms.; third, structural-equation modeling for microbial variables was constrained by small sample size. The SEM outputs should be treated as exploratory hypothesis rather than robust causal evidence. Larger sample sets are required in future work to validate these pathways.

Five consecutive field trials demonstrated that combined application of humic acid and chicken manure significantly increased average annual corn yield in Northeast China’s black soil regions. The five-year average yield increase exceeded 45% for both treatments. The humic acid treatment showed a higher interannual yield growth rate. It also exhibited greater long-term yield improvement potential. The chicken manure treatment showed lower yield variability and superior yield stability. Continuous organic material application significantly lowered soil pH over five years. However, its effects on bulk density, available nutrients, and organic carbon composition remained non-significant. Microbial community analysis revealed that organic materials did not substantially alter alpha diversity. However, they markedly modified community structure. This was achieved by enriching specific key microbial groups. Organic amendments altered microbial community composition without shifting overall microbial alpha-diversity, and soil pH was strongly correlated with microbial-community differentiation. Crucially, total N, P and K inputs were unbalanced among treatments, as organic materials supplied substantial additional macronutrients. Therefore yield differences cannot be causally attributed to organic-matter intrinsic properties, pH shifts or microbial-community changes alone.