The research team, led by Yuzhou Jiang and colleagues from Heilongjiang Bayi Agricultural University and the Chinese Academy of Agricultural Sciences, set out to address a persistent problem. The black soils of Northeast China are naturally alkaline, with pH values around 8.2 that restrict the availability of phosphorus and micronutrients, impair microbial functions, and cause year-to-year yield fluctuations under conventional chemical fertilization. As grain demand continues to grow across China, understanding how to stabilize productivity in these soils has become a matter of national food security.

The experiment compared three treatments in a randomized block design: a chemical fertilizer control (CK) receiving urea, diammonium phosphate, and potassium sulfate at recommended rates; a humic acid treatment (HA) adding 3,000 kilograms per hectare of humic acid extracted from weathered lignite; and a chicken manure treatment (JF) adding 1,500 kilograms per hectare of commercially composted poultry manure. All plots grew the same maize variety, Xianyu 335, under identical agronomic management. Over five harvests, the average yields under HA and JF reached 12.47 and 12.50 tonnes per hectare respectively, representing increases of 45.90 and 46.27 percent relative to the fertilizer-only control.

Yet the two organic amendments told distinctly different stories over time. Linear regression of yield trends revealed that the humic acid treatment exhibited a significantly steeper positive slope of 1.47 tonnes per hectare per year, with an R-squared of 0.58 and a P value of 0.001, compared with 0.84 tonnes per hectare per year for chicken manure. This suggests humic acid possesses superior long-term yield-building potential. Chicken manure, by contrast, delivered the lowest coefficient of variation at 17.64 percent, indicating better inter-annual stability, while humic acid showed the highest variability at 22.7 percent. The yield sustainability index ranged narrowly from 0.727 to 0.742 across all treatments, and the corrected yield trend index was lowest for humic acid at minus 49.93 percent, confirming a strong yield acceleration from the early to the later phases of the trial.

The soil chemistry results were equally revealing. After five years, both organic-amended treatments showed slightly but consistently lower soil pH than the control, with HA and JF plots measuring 8.02 and 8.01 respectively against 8.26 for CK. This represents a modest shift of approximately 0.25 pH units within the alkaline range. However, bulk density, alkaline-hydrolyzable nitrogen, available phosphorus, available potassium, and multiple organic carbon fractions, including total soil organic carbon, particulate organic carbon, mineral-associated organic carbon, easily oxidizable organic carbon, and water-soluble organic carbon, showed no significant differences among treatments. The researchers note that this carbonate-buffered calcareous soil has limited intrinsic capacity for such a small pH change to substantially modify phosphorus or micronutrient bioavailability on its own.

Metagenomic sequencing of rhizosphere soil, performed on samples collected at the final 2025 harvest, revealed a more nuanced microbial picture. Alpha diversity indices, including the Simpson, Chao1, Sobs, and Coverage indices, were highly similar across all treatments, with the Simpson index consistently at 0.99. This suggests that five years of organic amendment did not substantially alter overall microbial richness or evenness. However, principal component analysis showed distinct separation patterns: PC1 explained 14.62 percent of community variation and PC2 explained 13.86 percent, together accounting for 28.48 percent of total variation. The most pronounced differences in community structure appeared between the control and humic acid treatments, with chicken manure showing intermediate characteristics.

Taxonomic analysis uncovered treatment-specific enrichment patterns that hint at different ecological strategies. The humic acid treatment was descriptively associated with higher relative abundance of Luteitalea and Nocardioides, actinomycete-associated taxa linked to complex organic matter decomposition and rhizosphere substrate utilization. The chicken manure treatment enriched Bradyrhizobium, a genus strongly associated with nitrogen cycling, and Lysobacter, known for its antagonistic effects against other microorganisms. The control treatment maintained high relative abundance of Rubrobacter, the most abundant annotated genus across all plots. A Kruskal-Wallis H-test detected significant differences among treatments for ten genera, and LEfSe analysis further revealed distinct archaeal community compositions, though the researchers caution that several archaeal annotations, including thermophilic taxa biologically unexpected in temperate agricultural topsoil, likely reflect database annotation bias rather than genuine in-situ enrichment.

Correlation and ordination analyses pointed to soil pH as the primary environmental driver of microbial community differentiation. Redundancy analysis showed that RDA1 and RDA2 together explained 87.6 percent of community variation, with the pH vector carrying the longest arrow length. An exploratory structural equation model suggested positive associations between organic amendment treatment, reduced soil pH, and maize yield, with model coefficients of 0.92 for Luteitalea and minus 0.73 for Rubrobacter in relation to pH. However, the researchers are careful to emphasize that these modeled associations do not constitute mechanistic causal proof, given the small sample size and the confounding of nutrient inputs with organic material addition.

Indeed, the study’s most important caveat is its experimental design. Because the organic materials were applied at locally recommended rates rather than balanced for total nutrient content, the HA and JF treatments received substantial additional nitrogen, phosphorus, and potassium from organic sources. No equal-nutrient mineral-fertilizer control was included, meaning yield differences cannot be definitively attributed to the intrinsic properties of the organic materials, the modest pH shifts, or the observed microbial community changes alone. Interestingly, HA and JF received markedly different total nitrogen and phosphorus inputs from organic sources yet produced statistically indistinguishable multi-year average yields, an observation 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.

The practical implications remain significant for farmers across the Northeast Black Soil Region. Humic acid application should be prioritized when production goals emphasize long-term yield potential, while chicken manure proves more advantageous for achieving consistent annual yields and mitigating risk under climate variability. As the researchers conclude, both materials represent regionally adaptable organic soil enrichment practices, though their respective applications target different objectives. Future work with balanced nutrient controls and larger sample sets will be needed to disentangle the relative contributions of supplementary nutrients, pH modification, and microbial community restructuring to the remarkable yield gains documented over these five years.

Subject of Research: Long-term effects of combined chemical fertilizer and organic amendments on maize yield stability and soil microbial communities in alkaline black soils

Article Title: 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

Article References: Jiang, Y., Zhang, Y., Zhang, H., & Li, Y. (2026). 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. Frontiers in Agronomy, 8, Article 1948938. https://doi.org/10.3389/fagro.2026.1948938

Image Credits: AI Generated

DOI: 10.3389/fagro.2026.1948938

Keywords: maize yield, humic acid, chicken manure, alkaline soils, black soil region, soil microbiome, metagenomics, organic amendments, soil pH, yield stability, China agriculture, sustainable fertilization

The research team, led by Jie Zhou and Jianqiang He, built and compared five different machine learning models for predicting whether a late-spring frost event would strike apple orchards during the flowering period: Random Forest, Support Vector Machine, Extreme Gradient Boosting (XGBoost), a Multilayer Perceptron neural network, and a Long Short-Term Memory (LSTM) network. Crucially, they went beyond the meteorological data that most earlier studies relied on, adding spatial geographic variables including latitude, longitude, altitude, slope, and aspect to the predictor set. The study area covered China’s four major apple-growing regions: the Bohai Bay region, the Loess Plateau, the Southwestern Cool Highland, and Xinjiang. The Loess Plateau alone accounts for about 57.2 percent of national apple planting area, while the Bohai Bay region contributes roughly 26.9 percent.

The technical rationale for adding geography is rooted in the physics of frost formation. Frost at its core is a short-term low-temperature disaster in which plant tissues drop below 0 degrees Celsius, either through radiative cooling under clear, calm, high-pressure conditions or through advective frost when masses of cold air push temperatures below the frost point. The near-surface temperatures that determine whether buds, flowers, and young leaves freeze vary dramatically with terrain. Cold, dense air drains downslope at night and pools in narrow valleys and low-lying flats, while higher and more convex positions rarely freeze. Slope orientation matters too, because south-facing slopes in the Northern Hemisphere intercept more solar radiation and stay warmer than shaded north-facing slopes. A model blind to these spatial features is structurally incapable of capturing where frost will actually occur.

To train and test the models, the researchers compiled historical frost-damage records from 179 representative apple-growing sites in 2018 and 86 sites in 2020, drawing on the China Meteorological Disaster Management System, provincial meteorological bureaus, disaster yearbooks, and field investigations reported in the scientific literature. Frost severity was harmonized according to the Chinese Meteorological Industry Standard for flowering-stage frost damage in Fuji apples, which classifies injury as mild, moderate, or severe based on visible damage to petals, anthers, pistils, and ovaries and corresponding frost-injury rates. For the binary classification task, all severity levels were merged into a single frost-occurrence class, while non-frost samples were strictly defined as synchronous observations from the same sites and dates where no damage occurred. Predictor variables passed rigorous screening: average temperature was removed because its correlation with minimum temperature exceeded 0.7, and variance inflation factor tests confirmed that the remaining ten variables were free of problematic multicollinearity.

The results were striking. When the geographic variables were added to the meteorological set, every model improved. Overall accuracy rose to 0.900 for both Random Forest and XGBoost, 0.895 for the attention-enhanced LSTM, 0.882 for the Support Vector Machine, and 0.834 for the Multilayer Perceptron. The Random Forest model achieved the best discrimination overall, with an area under the receiver operating characteristic curve of 0.96 and a true positive rate of 0.914, meaning it correctly identified more than nine in ten actual frost events. XGBoost posted the lowest false alarm rate at 0.026, and the Support Vector Machine raised its true positive rate from 0.813 to 0.870 once geographic information was included. The team also introduced an LSTM variant augmented with a channel attention mechanism, which adaptively reweights input feature channels so the network emphasizes the most informative predictors. This LSTM-CAM model lifted overall accuracy from 0.861 to 0.895 and the true positive rate from 0.616 to 0.764 compared with a vanilla LSTM, demonstrating that attention-guided feature selection meaningfully improves multi-source data fusion.

Perhaps the most consequential methodological innovation was the application of SHAP, or Shapley Additive exPlanations, to open the black box. Derived from cooperative game theory, SHAP calculates the average marginal contribution of each input feature across all possible feature coalitions, yielding values that quantify exactly how much each variable pushed a given prediction toward or away from frost. Positive Shapley values indicate features that increased predicted frost probability, while negative values indicate suppressive effects. Crucially, SHAP delivers both global explanations, summarizing each variable’s average importance across all samples, and local explanations, revealing how individual predictions were formed. For farmers and foresters who must act on model outputs, this transparency transforms an opaque statistical verdict into an interpretable physical story.

What did the explanations reveal? Across every model, minimum temperature emerged as the single most influential predictor of late-spring frost, an intuitive but now quantified result. Solar radiation ranked second in the Random Forest and LSTM-CAM models: sunny days heat and charge the soil with energy that is then radiated away after sunset, so low daytime radiation signals a greater risk of nocturnal radiative cooling and frost. Wind speed also shaped the predictions, because strong winds mix warm air aloft down to the surface and disrupt the stagnant cold-air pooling that otherwise lets temperatures plummet in valleys and depressions. Among geographic factors, latitude was the leading driver in the Random Forest model, while altitude ranked highly in the Support Vector Machine and LSTM-CAM models, consistent with evidence that trees at higher elevations face rising frost exposure as warming advances spring phenology faster than frost dates retreat. Slope and aspect, being purely local terrain features, contributed less at regional scales but produced substantial local SHAP responses in specific conditions.

The framework’s real-world credibility was tested by asking the models to reproduce the severe 2018 frost disaster spatially across the four growing regions. All models captured the general pattern of the disaster, but with telling differences in spatial realism. The Random Forest model showed the highest spatial agreement with observed frost-damage rates on apple central flowers, correctly delineating the high-risk belts across the Loess Plateau and parts of the Bohai Bay region while assigning low probabilities to the comparatively unscathed Xinjiang and Southwestern Cool Highland regions. The Support Vector Machine and XGBoost models produced more fragmented high-risk zones and overestimated frost in areas like parts of Liaoning Province. The LSTM-CAM model stood out for suppressing spurious high-probability signals in Xinjiang, where observed impacts were mild, illustrating how attention mechanisms filter weakly relevant or noisy inputs. In an independent out-of-sample validation using localized frost events from 2012 and 2013 that the models had never seen, the LSTM-CAM architecture achieved the highest AUC values of 0.81 and 0.84 and true positive rates of 0.67 and 0.70, demonstrating stronger cross-year generalization than the tree ensembles when frost events were spatially sparse.

The authors are candid about the framework’s limits. All training samples came from the flowering period, so phenological susceptibility was implicitly controlled by the sampling design rather than modeled explicitly; incorporating observed or simulated flowering dates could sharpen the biological timing of risk. High-quality, spatially continuous frost-damage records remain scarce and concentrated in a few extreme years, constraining the models’ ability to capture interannual variability. Site-based observations also measure air temperature at standard heights, which can diverge substantially from the tissue-level temperatures at which flowers actually freeze, and frost thresholds calibrated for one region may mislead in another. The researchers argue that region-specific thresholds and multi-source long-term datasets are essential next steps.

Even with those caveats, the study delivers a scalable, transparent early-warning tool at a moment when climate change is reshaping frost risk worldwide. Warmer winters advance bud break and flowering, while cold intrusions still arrive, and the resulting collision between sensitive phenological stages and subzero air threatens yields in every major temperate fruit region. By proving that geographic context materially improves frost prediction, and by showing that explainable AI can reveal not just what a model predicts but why, the work offers orchardists and policymakers a defensible basis for targeted frost protection, from wind machines that break up cold-air pooling to siting decisions for new plantings. The next devastating frost may still be unavoidable, but with interpretable machine learning watching the weather and the terrain, it need not be a surprise.

Subject of Research: Explainable machine learning prediction of late-spring frost risks in apple production using meteorological and geospatial features

Article Title: Integrating explainable machine learning models with geospatial features to enhance the prediction of late-spring frost risks in apple production

Article References: Zhou, J., Ru, X., He, Z., Gong, K., Wang, J., Wang, Z., Li, M., Feng, H., Yu, Q., & He, J. (2026). Integrating explainable machine learning models with geospatial features to enhance the prediction of late-spring frost risks in apple production. Artificial Intelligence in Agriculture. https://doi.org/10.1016/j.aiia.2026.08.017

Image Credits: AI Generated

DOI: 10.1016/j.aiia.2026.08.017

Keywords: late-spring frost, apple production, machine learning, explainable AI, SHAP, Random Forest, LSTM, attention mechanism, geospatial features, frost prediction, China agriculture, climate risk