The research applies the Push-Pull-Mooring framework, a model originally developed to explain why consumers switch service providers, to a setting far removed from mobile phone plans or banking apps: capital-intensive, perennial horticulture in a mountain economy. The framework posits that a person’s decision to abandon an existing practice for a new one is driven by three interacting sets of forces. Push factors are the dissatisfactions that make the status quo untenable. Pull factors are the attractions that make the alternative appealing. Mooring factors are the anchors, both personal and structural, that hold people in place or ease their transition. What makes this application particularly compelling is that switching apple production systems is not a casual decision; an orchard represents a multi-year, often multi-generational investment, and a wrong choice can burden a family for a decade or more.

To capture these dynamics empirically, the team designed a structured questionnaire and administered it to 275 apple growers across Jammu and Kashmir, one of India’s most important apple-producing regions. The study employed a cross-sectional research design with two-stage purposive and judgmental sampling, a strategy appropriate for reaching growers who vary in their exposure to and experience with high-density systems. The researchers then analysed the data using Partial Least Squares Structural Equation Modelling, or PLS-SEM, a statistical technique well suited to testing complex networks of relationships between latent psychological constructs, including the mediating pathways the model proposed. All computations were carried out in Smart PLS, allowing the team to estimate both the direct effects of push, pull and mooring factors and their indirect effects transmitted through behavioural intention.

The results carry a striking headline finding: the push factors, the dissatisfactions with conventional orchards, exert the most significant influence on growers’ intention to switch. These dissatisfactions are concrete and accumulating. Traditional orchards deliver low yields per unit of land, produce fruit of poorer and less uniform quality, demand intensive labour across large canopies, impose long gestation periods before trees bear marketable fruit, and expose growers to escalating risks posed by a changing climate. In other words, the study suggests that the engine of agricultural modernisation in this region is not primarily the allure of the new, but the growing untenability of the old. Farmers are being pushed out of their traditional systems by pain, not merely lured by promise.

That said, the pull factors matter considerably as well. The expected yield gains of high-density systems, the prospect of premium prices for higher-quality fruit, labour efficiency from smaller, more manageable trees, early bearing that shortens the wait for returns, and supportive policy incentives all significantly increase growers’ adoption intention. High-density orchards, which plant dwarfing rootstocks at far closer spacing than traditional seedling trees, can transform the economics of apple farming: more tonnes per hectare, earlier first harvests, easier spraying and pruning, and fruit that meets the quality standards demanded by modern markets. The study’s confirmation that these benefits register strongly in growers’ minds underscores that the technology itself, when well communicated, carries genuine persuasive power.

Between the push and the pull sits the mooring, and here the study reveals the friction in the system. Behavioural inertia, the deep-rooted habit of farming as one’s parents and neighbours farmed, conditions the switching process. So do more tangible constraints: access to inputs such as quality planting material and fertilisers, access to training in the new agronomic techniques that high-density systems demand, and perceptions of financial risk, which loom large given the substantial upfront investment these orchards require. The expected yield gains themselves also condition the switching process, acting as a bridge between aspiration and action. These findings suggest that even when farmers are dissatisfied with the old and enthusiastic about the new, they may remain anchored unless the practical supports for transition are in place.

One of the study’s most important structural findings concerns the role of behavioural intention as a mediator. Push, pull and mooring forces do not act on actual behaviour directly; instead, their effects flow through intention, which in turn is a good predictor of actual use behaviour. This mediating architecture has practical implications. It means that interventions aimed at changing farmer behaviour, whether subsidies, demonstration plots, or extension programmes, work by reshaping intentions, and that understanding and measuring intention is a legitimate and informative proxy for tracking the progress of agricultural modernisation. It also means that policies which address only one side of the equation, for instance by offering incentives without relieving the dissatisfactions of the old system, or by highlighting benefits without removing mooring constraints, may underperform.

In terms of explanatory power, the model accounts for 51 per cent of the variation in behavioural intention and 23 per cent of the variation in use behaviour. The authors characterise this as moderate explanatory relevance for intention and modest relevance for actual behaviour, an honest assessment that reflects the realities of studying human decisions in complex, real-world settings. Intention, after all, is easier to predict than action, particularly when action depends on credit availability, weather, market fluctuations and family circumstances that no psychological model can fully capture. Still, explaining half of the variance in a behavioural intention within a heterogeneous farming population is a substantial achievement, and the 23 per cent explained for use behaviour provides a meaningful foothold for policy design.

The context of the study gives its findings particular weight. The Indian Himalayan region faces a convergence of pressures: shrinking landholdings, ageing orchards, water stress, erratic weather and the steady encroachment of climate change on a crop that defines the regional economy. Apple cultivation in Jammu and Kashmir supports hundreds of thousands of households, and the transition to high-density systems is widely seen as central to keeping the industry viable. Yet the transition is capital-intensive in a region where credit access can be uneven, and it demands new knowledge, from canopy management to rootstock selection, that traditional extension systems have struggled to deliver at scale. By quantifying which forces actually move growers, the study offers mountain economies a diagnostic tool for accelerating modernisation without wasting resources on interventions that behavioural evidence does not support.

For policymakers, the message is a three-part prescription grounded in the data. First, amplify and validate the push: help growers accurately assess the declining productivity and rising risk of their conventional orchards, so that dissatisfaction is grounded in evidence rather than rumour. Second, strengthen the pull: ensure that the demonstrated yield gains, premium market access, early returns and policy incentives of high-density systems are visible, credible and locally attested. Third, and perhaps most critically, cut the moorings: expand access to affordable finance, certified planting material and hands-on training, and design programmes that acknowledge and work with behavioural inertia rather than ignoring it. As the authors note, their findings provide context-specific insights for Himalayan apple-growing regions, but the framework they deploy, and the lesson that switching is a system of forces rather than a single decision, should resonate with anyone working to modernise agriculture in mountain economies around the world.

Subject of Research: Behavioural drivers of apple growers' adoption of high-density sustainable farming systems in the Indian Himalayan region

Article Title: Exploring apple growers’ switching behaviour towards high-density sustainable farming using the push-pull-mooring framework

Article References: Amin, I. U., Bhat, S. A., Seher, N., & Shafi, S. I. (2026). Exploring apple growers’ switching behaviour towards high-density sustainable farming using the push-pull-mooring framework. Discover Sustainability. https://doi.org/10.1007/s43621-026-04856-5

Image Credits: AI Generated

DOI: 10.1007/s43621-026-04856-5

Keywords: high-density apple systems, push-pull-mooring framework, behavioural intention, technology adoption, agricultural modernisation, mountain agriculture, PLS-SEM, Jammu and Kashmir, sustainable farming, apple growers, switching behaviour, climate risk

The research team, led by Maha Elbana of Beni-Suef University in Egypt, alongside collaborators from the China Meteorological Administration, the National Authority for Remote Sensing and Space Sciences, and the International Center for Agricultural Research in the Dry Areas, conducted their experiment at the Arid Lands Agricultural Studies and Research Institute of Ain Shams University. Their goal was ambitious: to simultaneously track water use, energy demand, nitrogen efficiency, fruit quality, and crop yield under conditions where a shallow water table sits close enough to the root zone to buffer the soil’s moisture supply.

The experimental setup was carefully controlled. Tomato plants of the variety Rich Zoan K186 were grown in clay loam soil—composed of 34.4 percent clay, 35.9 percent silt, and 29.7 percent sand—across two growing seasons: Spring 2023 and Autumn 2023-2024. A pressurized drip irrigation system, governed by daily soil-moisture readings from time-domain reflectometry probes, ensured that every plot received water only when root-zone moisture dropped to 50 percent of available water. This uniform irrigation protocol meant that differences among treatments could not be attributed to water stress; instead, the researchers manipulated two factors: nitrogen rate and biochar application.

Six treatment combinations were tested. Three nitrogen rates—431, 345, and 215 kilograms of nitrogen per hectare, corresponding to 100, 80, and 50 percent of the local farmer practice—were crossed with two biochar levels: zero or 35 tonnes per hectare. The biochar, derived from corn cobs, was incorporated into the top 30 centimeters of soil one week before transplanting. Each plot was hydraulically isolated with trenches to minimize lateral water movement, and the experiment was arranged in a randomized complete block design with three replicates.

The most striking result was what did not happen. Marketable fresh yield showed no statistically significant differences among any of the six treatment combinations, ranging from 81.4 to 108.6 tonnes per hectare in pooled treatment means. In a hydrologically buffered system—where shallow groundwater and high soil-water storage in fine-textured soil maintain root-zone moisture near a defined threshold—the crop simply did not respond to changes in nitrogen input with the yield differences that conventional agronomy might predict. The researchers’ null hypothesis for yield was not rejected.

But water productivity told a more nuanced story. A significant Season × Nitrogen-rate interaction emerged (F2,8.52 = 5.50, p = 0.029), indicating that the response of water productivity to nitrogen input varied between the two growing seasons. However, when the researchers applied Tukey-adjusted pairwise comparisons within each season, no individual nitrogen rate proved statistically superior to another. The pattern suggests that seasonal environmental conditions—temperature, radiation, evaporative demand—modulated how efficiently the crop converted water into yield under different nitrogen regimes, but not in a way that consistently favored one rate over another.

The energy accounting revealed the experiment’s most consequential finding. Fertilizer-associated nitrogen energy—calculated using an embodied energy coefficient of 60.6 megajoules per kilogram of nitrogen, reflecting the intensive industrial process of mineral nitrogen production—dominated the total accounted energy budget. At the full nitrogen rate, fertilizer-associated energy reached 26,143 megajoules per hectare, compared with just 1,270 to 1,831 megajoules per hectare for irrigation pumping. In other words, the energy embedded in nitrogen fertilizer was roughly 7 to 20 times greater than the energy required to pump irrigation water. Cutting nitrogen to 80 percent reduced fertilizer-associated energy by approximately 20 percent; cutting it to 50 percent reduced it by half.

Total accounted energy—combining irrigation pumping and fertilizer-associated nitrogen energy—ranged from approximately 14,430 megajoules per hectare under the 50 percent nitrogen treatments to over 27,500 megajoules per hectare under the full rate. Because irrigation energy remained statistically unchanged across treatments (p > 0.146), the nitrogen rate was the primary determinant of the overall energy footprint. This finding carries significant implications for arid-region agriculture, where mineral nitrogen production represents a substantial share of greenhouse gas emissions and where reducing unnecessary fertilizer inputs simultaneously addresses soil acidification, nitrate leaching, and climate burdens.

The integrated performance profiles, visualized through directionally normalized radar plots and standardized heatmaps, exposed trade-offs that yield alone could never reveal. Treatment W8—no biochar, 80 percent nitrogen—produced the highest pooled numerical yield at 108,583 kilograms per hectare while reducing fertilizer-associated energy by roughly 20 percent compared with the full rate. Meanwhile, treatment B5—biochar, 50 percent nitrogen—achieved the highest pooled water productivity at 39.59 kilograms per cubic meter, the highest energy productivity at 6.87 kilograms per megajoule, and a partial factor productivity of nitrogen of 459.09 kilograms of yield per kilogram of nitrogen applied. These ratio-based indicators, the researchers caution, increase partly because of smaller denominators and should not be interpreted as proof of agronomic superiority on their own.

Fruit quality responses were governed primarily by season rather than by management. Fruit acidity, nitrogen concentration, hardness, and nitrate-nitrogen levels all showed significant seasonal effects, while biochar and nitrogen-rate effects were restricted to specific variables and treatment contexts. The only statistically significant adjusted pairwise difference among nitrogen rates occurred for fruit acidity in spring under biochar, where the full nitrogen rate produced higher acidity than the 80 percent rate. Reducing nitrogen input did not produce a generalized deterioration in fruit quality, supporting the notion that moderate nitrogen reduction can maintain both yield and fruit composition under well-managed drip irrigation.

The broader implications extend beyond a single field in Egypt. The study demonstrates that in hydrologically buffered systems—where shallow groundwater, fine soil texture, and sensor-based irrigation conspire to equalize water availability across treatments—yield becomes a blunt instrument for evaluating management performance. Integrated indicators of water productivity, energy productivity, and nitrogen-use efficiency reveal contrasts that yield conceals. The researchers emphasize that their energy accounting represents a partial operational boundary, excluding biochar production, irrigation infrastructure, and post-harvest operations, and that future work should quantify nitrogen uptake, expand the life-cycle boundary, and test whether these trade-offs persist across different soil textures and longer timeframes. For now, the message is clear: two fields that look identical by the harvest scale may be worlds apart in the energy they consume and the resources they demand.

Subject of Research: Integrated water, energy, and nitrogen performance of drip-irrigated tomato production under varying nitrogen rates and biochar application in shallow groundwater conditions

Article Title: Integrated water–energy–nitrogen performance under nitrogen and biochar treatments in shallow groundwater tomato production

Article References: Elbana, M., Wang, L., El-Shirbeny, M. A., Gamal, R., Nangia, V., & Abou Hadid, A. F. (2026). Integrated water–energy–nitrogen performance under nitrogen and biochar treatments in shallow groundwater tomato production. Frontiers in Agronomy, 8, Article 1922454. https://doi.org/10.3389/fagro.2026.1922454

Image Credits: AI Generated

DOI: 10.3389/fagro.2026.1922454

Keywords: tomato, nitrogen fertilizer, biochar, drip irrigation, water productivity, energy productivity, shallow groundwater, arid agriculture, soil moisture, fertilizer energy, Egypt, sustainable farming

The framework rests on a three-tier architecture that moves data from the field to the farmer’s phone. At the bottom sits an ultra-low-power sensor node built around the MSP430G2553, a 16-bit microcontroller chosen specifically for its frugal energy appetite. The node carries a DS18B20 soil temperature sensor and a capacitive soil moisture sensor, both connected over the I2C protocol and inserted 10 centimeters into the soil. A Li-Ion battery powers the node, recharged by a solar panel through a TP4056 charger, which makes the data collection process self-sustaining in the field. An M95 GSM modem links the node to the internet, and readings are stored in the microcontroller’s EPROM before being transmitted to the next tier every 30 minutes over Wi-Fi.

The middle tier is where the system departs from conventional designs. Instead of shipping raw sensor data straight to a remote cloud, a Raspberry Pi edge device with 16 GB of RAM performs the first round of processing locally. A Python script converts the raw microcontroller data into a csv file stamped with date and time, replaces missing values with the mean of the respective attribute, removes outliers, and generates a cleaned dataset. Then comes the privacy step: before anything leaves the farm, the edge device adds Gaussian noise to the pre-processed data, a technique known as Gaussian differential privacy, or GDP. The noisy csv file, along with metadata about the soil type, whether sandy, loamy, or clayey, is uploaded to the cloud just once per day.

Gaussian differential privacy is the system’s answer to a growing anxiety in agricultural data science. Sending raw field data to a third-party cloud server exposes farmers to potential surveillance and misuse, but full encryption adds computational overhead that small edge devices can ill afford. GDP offers a lightweight middle path: by injecting calibrated Gaussian noise scaled to the sensitivity of the data, it mathematically limits what any observer can infer about individual records while preserving the overall statistical pattern of the dataset. The researchers added noise within a scale of (0, 1) and found that the correlation between soil moisture and soil temperature remained essentially unchanged, closely matching the correlation computed on the original, un-noised data. The trade-off is real but modest: accuracy on the GDP-protected dataset dropped only slightly, from 99.5 percent on raw data to 99 percent, and latency rose by fractions of a second.

Once the privacy-protected data reaches the cloud, hosted on Google Cloud Platform and connected through a NodeJS client using the Message Queuing Telemetry Transport protocol, the analytical heavy lifting begins. The team chose ensemble learning over single-classifier machine learning, exploiting all three of its major flavors. Bagging trains models independently on random subsets of the data and combines their outputs by voting or averaging, implemented here with random forest. Boosting trains weak learners sequentially, each correcting the errors of its predecessor, implemented with extreme gradient boosting, or XGBoost. Stacking trains different model types and feeds their predictions into a meta model, in this case logistic regression, which learns the optimal combination of the base models’ outputs.

The experimental results are striking. Using a dataset of 10,000 samples split 80-20 into training and test sets, and validated with 5-fold cross-validation, all three ensemble models achieved 99 percent accuracy, precision, recall, and F1-score. Boosting with XGBoost matched the latency and energy consumption of single classifiers like KNN and decision tree, at 0.45 seconds and 315 joules, while bagging with random forest came in at 1.77 seconds and 1,250 joules. Stacking was the most expensive of the three at 9.21 seconds and 6,450 joules. The deep learning baselines, LSTM and GRU networks, matched the 99 percent accuracy but at a punishing cost: 12.38 seconds and 8,700 joules for LSTM, and 10.33 seconds and 7,240 joules for GRU. In other words, the ensemble approach delivered deep-learning-level accuracy with a fraction of the computational burden, a decisive advantage for deployment in resource-constrained rural settings.

The system also anticipates failure. Sensors in agricultural fields break, get damaged by weather, or simply are not installed in the first place, and in developing regions buying multiple sensor types may not be economically feasible for farmers. To cope, the researchers built regression models that predict soil temperature from soil moisture data and vice versa, so that irrigation decisions can proceed even when only one sensor is working. Again the ensemble methods outperformed the KNN and decision tree baselines: for soil temperature prediction, stacking achieved the lowest mean absolute error at 0.81 and the best coefficient of determination at 0.48, while for soil moisture prediction stacking again led with an MAE of 7.91 and an R-squared of 0.45. Interestingly, soil temperature could be predicted more accurately from moisture data than the reverse.

Robustness was tested beyond the team’s own fields. The framework was evaluated on two publicly available datasets, one with 3,000 samples covering soil moisture, air temperature, and air humidity labeled with motor on/off decisions, and another with 200 samples covering crop type, moisture, and temperature labeled with pump on/off decisions. After applying the same Gaussian noise, the ensemble models again reached 99 to 99.5 percent accuracy, with XGBoost once more combining top-tier accuracy with the lowest latency and energy consumption, 0.16 seconds and 192 joules on the larger dataset. Compared against published irrigation systems, which reported accuracies ranging from 87.5 percent for SVM-based approaches to 99.96 percent for a standalone XGBoost model, ECIID’s 99 percent figure stands out for being achieved while simultaneously incorporating edge processing, latency measurement, soil temperature prediction, and privacy protection, features the authors’ comparison table shows no prior system combined.

The researchers also subjected their framework to A/B testing, reporting Z-scores above 2.58 and P-values of 0 for all three ensemble models on their dataset, which they interpret as evidence of statistical robustness and high confidence. Data collection ran for 90 days across three regions with different soil types, and the labeling rules were soil-specific: for loamy soil, irrigation is triggered when temperature falls to a threshold and moisture reaches 40 percent; for sandy soil, which drains faster, the moisture threshold is 50 percent; for clayey soil, which retains water, the trigger is above a temperature threshold with 40 percent moisture. These rules reflect the agronomic reality that different seasons and soil textures demand different moisture levels for optimal crop growth.

The final piece is practical: an Android app that lets farmers log in, view real-time soil temperature and moisture data, check the irrigation decision, and download collected and predicted data as a csv file, all from anywhere. Because GDP is applied before transmission, the app shows data that preserves the pattern of the original dataset but not the exact values, a limitation the authors acknowledge openly. Looking ahead, the team plans to address scenarios where both sensors fail simultaneously or extreme weather disrupts monitoring, and to develop a split learning-based irrigation model for precision agriculture. For now, ECIID offers a template for how edge computing, differential privacy, and ensemble learning can converge to make every drop of irrigation water count, and to count it without giving away the farm’s data.

Subject of Research: IoT and edge computing-based soil moisture monitoring with ensemble learning for sustainable irrigation decision making

Article Title: Ensemble learning-based soil moisture monitoring and sustainable irrigation management for Internet of Agricultural Things

Article References: Bera, S., Dey, T., Sengupta, A., Mukherjee, A., & De, D. (2026). Ensemble learning-based soil moisture monitoring and sustainable irrigation management for Internet of Agricultural Things. Frontiers in The Internet of Things, 5, Article 1870523. https://doi.org/10.3389/friot.2026.1870523

Image Credits: AI Generated

DOI: 10.3389/friot.2026.1870523

Keywords: smart irrigation, soil moisture, ensemble learning, Internet of Things, edge computing, Gaussian differential privacy, XGBoost, random forest, precision agriculture, machine learning, data privacy, sustainable farming

The stakes are far from trivial. Andalusia accounts for 21% of Spain’s agricultural area, and agriculture there is responsible for 7,005.07 Gg of CO2 equivalent—29.2% of the region’s diffuse-sector emissions, which include transport, agriculture and waste. Those emissions arise mainly from nitrogen fertilisers and fossil fuels, and conventional tillage, the ploughing-based system still common across the Andalusian countryside, burns more diesel than any system that minimises soil disturbance. Conventional ploughing also oxidises soil organic carbon, eroding natural fertility and fertiliser-use efficiency over time.

Conservation Agriculture rests on three interlinked principles: minimal soil disturbance, permanent soil cover with crop residues, and diversified crop rotations. Precision Agriculture, by contrast, envisages a more rational use of inputs by eliminating the overlaps between machine passes—areas where a sprayer or spreader covers ground already treated on the previous pass. The research team hypothesised that combining the two, specifically no-tillage with residue retention plus GPS guidance assistance, would outperform conventional tillage on both energy and carbon metrics. To test this, they monitored 90 hectares across three commercial farms in Córdoba, Carmona and Las Cabezas de San Juan from the 2009/10 to the 2012/13 growing seasons, following the classic Andalusian rotation of cereal, sunflower and legumes.

The instrumentation was meticulous. Three tractors—a John Deere 6420, a Fendt Vario 712 and a New Holland TM 120—were fitted with potentiometers to detect the rear-hitch position, GPS receivers to calculate task duration, worked area and overlap, and flow meters to measure instantaneous fuel consumption. A data logger recorded UTM position, heading, speed, fuel use and hydraulic-lift position every two seconds, transmitting data daily via GPRS to a remote computer. The tractors used for guidance treatments were equipped with a Trimble AgGPS EZ-Guide 500 for spraying and fertiliser application and an EZ-Steer assisted steering system for sowing. Yields were sampled in ten homogeneous 0.5-hectare grids per plot using a one-square-metre frame, giving a robust picture of production under each system.

The energy analysis followed the framework proposed by the International Federation of Institutes for Advanced Study, distinguishing direct-use energy—fuel burned in the field, converted at 36.8 MJ per litre—from indirect-use energy embedded in machinery manufacture and maintenance, seeds, fertilisers and plant protection products, calculated on a cradle-to-gate basis. Two headline indicators emerged: energy efficiency, the ratio of energy contained in the harvested grain to the energy required to produce it, and energy productivity, the kilograms of food produced per megajoule of energy supplied. Carbon emissions were derived from fuel consumption using a diesel conversion factor of 0.074 kg CO2 equivalent per MJ, with emission factors for production inputs drawn from the Spanish industrial energy mix of 2009 to 2012.

The fuel data told the clearest story. Primary tillage was the most fuel-intensive operation, with the mouldboard plough consuming 22.5 litres per hectare and the chisel 14.1 litres per hectare. Under conventional tillage, the disc harrow recorded the highest overlap at 35.1%, and a total of 136 tillage-related passes—mouldboard plough, chisel, disk harrow, semi-chisel, scarifier, cultivator, baler and roller—were conducted that no-tillage systems simply eliminate. The only notable increase under no-tillage was in sprayer passes, 29 more than under conventional tillage, because weed control shifts from mechanical cultivation to herbicides. Even so, the arithmetic was decisive: average fuel consumption fell by 53% overall, with reductions of 55.7% for legumes, 52.9% for sunflower and 51.6% for wheat, all statistically significant.

Energy consumption followed the same pattern. Wheat was the most energy-demanding crop, consuming 17.54 GJ per hectare under the combined system and 19.92 GJ per hectare under conventional tillage, largely because of its high nitrogen fertiliser requirements. Sunflower required the least energy, at 2.75 GJ per hectare under the combined system versus 3.73 GJ per hectare under conventional tillage. For wheat, the elimination of tillage accounted for 35.8% of the absolute energy saving, while the remaining 64.2% came mainly from fertiliser savings enabled by reduced overlap under guidance assistance. For sunflower and legumes, where fertiliser use was minimal, more than 90% of the saving came from suppressing tillage. The one production factor where the combined system consumed more energy was plant protection products, averaging 45% higher than conventional tillage, driven by herbicide use—though the researchers note that reduced runoff and erosion under no-tillage limits non-point pollution, and continuous no-till increases soil organic matter, enhancing herbicide degradation in soil.

Crucially, none of these savings came at the cost of production. Yields did not differ significantly between systems: wheat averaged 2.39 Mg per hectare under the combined system versus 2.24 Mg per hectare under conventional tillage, sunflower 1.08 versus 1.10, and legumes 1.06 versus 1.08. Energy efficiency and energy productivity improved under the combined system for all crops, with sunflower showing the only statistically significant gains—a 33.5% increase in energy efficiency and 34.5% in energy productivity. Sunflower was the most efficient crop overall, with efficiency and productivity values roughly twice those of legumes and three times those of wheat, despite producing less total energy than the cereal.

The carbon balance mirrored the energy results. Greenhouse gas emissions fell by 26.4% for sunflower, 18.4% for legumes and 12.0% for wheat under the combined system, all statistically significant. Carbon use efficiency and carbon use productivity—kilograms of carbon harvested per kilogram of carbon emitted, and kilograms of food produced per kilogram of CO2 equivalent emitted—were 15 to 35% higher under the combined system. The study deliberately excluded soil carbon sequestration and soil respiration, focusing solely on emissions from field operations and production inputs; the authors note that when sequestration is considered, the carbon balance becomes even more favourable to Conservation Agriculture, since minimally disturbed soils with retained residues accumulate organic carbon while conventional tillage accelerates its oxidative breakdown.

The implications reach well beyond Andalusia. Under the European Climate Law, the EU mandates a net greenhouse gas reduction of at least 55% by 2030 relative to 1990 levels and climate neutrality by 2050, and agriculture is under growing pressure to contribute. The researchers argue that their simple, replicable methodological approach—built on real, remotely monitored field data—can help farmers and trainers identify where energy is best invested in relation to production, crop by crop and system by system. In a Mediterranean region where climate change is expected to hit particularly hard, the message is clear: the plough is no longer the default. A tractor that never tills and never strays from its GPS line can grow the same food with half the fuel and a markedly lighter carbon footprint.

Subject of Research: Agricultural Science

Article Title: Improving energy efficiency and carbon balance through conservation and precision agriculture with respect to conventional agriculture: a methodological approach

Article References: Márquez-García, F., Repullo-Ruibérriz de Torres, M. Á., Basch, G., & González-Sánchez, E. J. (2026). Improving energy efficiency and carbon balance through conservation and precision agriculture with respect to conventional agriculture: a methodological approach. Frontiers in Agronomy, 8, Article 1916869. https://doi.org/10.3389/fagro.2026.1916869

Image Credits: AI Generated

DOI: 10.3389/fagro.2026.1916869

Keywords: Conservation Agriculture, Precision Agriculture, No-Tillage, Carbon Balance, Energy Efficiency, GPS Guidance, Sustainable Farming, Climate-Smart Agronomy, Improving, energy, efficiency, carbon

The problem the researchers set out to tackle is more insidious than either stress alone. Salinity does not simply dehydrate plants; it also changes soil chemistry in ways that make heavy metals more mobile and more available for uptake. Earlier work cited by the team, including studies on cadmium mobility in saline soils, has shown that sodium salts can displace metal ions from soil particles, releasing them into the soil solution where roots readily absorb them. In the experimental soil used in this study, which had an electrical conductivity of 9.06 decisiemens per meter, wheat plants accumulated substantially more sodium, copper, lead and cadmium than plants grown under normal conditions.

The consequences for the plants were severe. Across two growing seasons, salinity combined with heavy metal contamination raised leaf sodium concentrations by roughly 50 to 53 percent, while leaf contents of copper rose by about 24 percent, lead by 66 to 87 percent and cadmium by 36 to 41 percent. The contaminated plants also accumulated more nitrate and nitrite, compounds that are undesirable in grain destined for human consumption. The physiological toll was reflected in the harvest: straw yields fell by about 30 to 35 percent and grain yields by nearly 32 percent compared with unstressed controls.

At the biochemical level, the stressed plants were fighting a losing battle against reactive oxygen species. Salinity and heavy metals both trigger the overproduction of these destructive molecules, which damage membranes, proteins and DNA. The plants responded by ramping up their enzymatic antioxidant defenses—superoxide dismutase, catalase and guaiacol peroxidase—and by elevating their pools of glutathione, a tripeptide that sits at the heart of the cellular redox system. Yet protein content declined, and the levels of essential macronutrients such as nitrogen, phosphorus and calcium, along with micronutrients including iron, manganese and zinc, all dropped. The plants were essentially starving while simultaneously being poisoned.

The intervention tested by the team was elegantly simple. Seeds were primed in either water, a 3 percent moringa leaf extract, a 1 millimolar glutathione solution, or a combination, and the same solutions were then applied as foliar sprays as the plants developed. Moringa oleifera, often called the miracle tree, is rich in antioxidants, vitamins, minerals and growth-promoting substances, and its leaf extract has previously been shown to alleviate salt and cadmium stress in several crops. Glutathione, meanwhile, is the plant’s own master antioxidant, participating directly in the detoxification of reactive oxygen species and in the chelation and sequestration of heavy metals.

The results of the sequenced application were striking. Plants that received both moringa leaf extract and glutathione, as seed priming followed by foliar spray, showed increases in superoxide dismutase activity of roughly 45 to 49 percent and in guaiacol peroxidase activity of 51 to 56 percent compared with stressed, untreated plants. Total protein content climbed by about 32 to 33 percent, and glutathione levels themselves surged by 60 to 70 percent, suggesting that the treatment primed the plants’ internal antioxidant machinery rather than merely supplementing it from outside.

Just as importantly, the treatment restored nutritional balance. Leaf concentrations of nitrogen rose by 28 to 51 percent, phosphorus by nearly 48 to 50 percent, potassium by 20 to 26 percent and calcium by 33 to 53 percent. Micronutrients followed the same pattern: leaf iron increased by 33 to 51 percent, zinc by 48 to 54 percent and manganese by about 36 percent, with similarly large gains in the grain. Because iron, zinc and manganese deficiencies affect billions of people worldwide, the finding that this treatment can biofortify wheat grain while detoxifying it is among the most consequential aspects of the work.

At the same time, the toxic load in the plants fell sharply. Leaf cadmium declined by 47 to 50 percent and lead by 48 to 49 percent, while grain cadmium dropped by 29 to 30 percent, lead by 47 to 54 percent and copper by about 26 percent. Leaf sodium, the principal driver of salt toxicity, fell by 40 to 45 percent, and grain nitrate and nitrite contents decreased by 47 to 54 percent and 54 to 61 percent respectively. The combined effect was a substantial recovery in productivity, with treated plants producing markedly higher straw and grain yields than their untreated, stressed counterparts.

The mechanism, as the authors describe it, is a coordinated restoration of homeostasis. By boosting the antioxidant defense system, the treatment limits the oxidative damage that heavy metals and sodium inflict on cells. By improving nutrient uptake, it counteracts the competitive interference that sodium causes with potassium, calcium and other essential elements. And by reducing the accumulation of toxic ions, it frees the plant’s metabolic resources for growth and grain filling rather than repair. The sequenced approach—priming seeds to establish defenses early, then reinforcing them with foliar sprays during critical growth stages—appears to be key, consistent with the team’s earlier work showing that sequenced glutathione and biostimulant applications improve wheat’s adaptation to salinity.

For agriculture in arid and semi-arid regions, where saline and contaminated soils are spreading, the implications are considerable. Both moringa leaf extract and glutathione are inexpensive, environmentally benign and easy to apply, making the strategy accessible to smallholder farmers who cannot afford soil remediation or expensive genetic solutions. The study, funded by the Deanship of Scientific Research at King Abdulaziz University, demonstrates that crop protection on degraded land need not begin in the soil itself; it can begin in the seed and the leaf. If field-scale trials confirm these two-season results, a tree extract and a common cellular antioxidant could become standard tools for growing safer, more nutritious wheat on some of the world’s most compromised farmland.

Subject of Research: Alleviation of combined salinity and heavy metal stress in wheat using moringa leaf extract and glutathione

Article Title: Glutathione and moringa leaf extract improve wheat tolerance to heavy metal-contaminated saline soil by regulation of antioxidant defense system and nutrient balance

Article References: Alamri, A. M., Rehman, H. U., Khalid, F., Alharby, H. F., Majrashi, A., Abdulmajeed, A. M., & Rady, M. M. (2026). Glutathione and moringa leaf extract improve wheat tolerance to heavy metal-contaminated saline soil by regulation of antioxidant defense system and nutrient balance. Plant Cell Reports, 45(11), Article 329. https://doi.org/10.1007/s00299-026-04010-5

Image Credits: AI Generated

DOI: 10.1007/s00299-026-04010-5

Keywords: wheat, salinity stress, heavy metals, glutathione, moringa leaf extract, antioxidant defense, seed priming, foliar spray, nutrient balance, grain quality, cadmium, soil contamination

The study, published in the International Journal of Machine Learning and Cybernetics, was led by Anish Kumar Vishwakarma and Rakesh Ranjan, who set out to fix a structural weakness in how deep learning has been applied to agricultural image classification. Most existing pest recognition systems are built as multi-stage pipelines: one model extracts features from an image, and a separate classifier makes the final decision. This division of labor sounds tidy, but it carries real costs. Each stage adds computational overhead, and because the feature extractor is not trained jointly with the classifier, the learned representations can miss the subtle visual cues that separate one pest species from another. For insects that differ only in the pattern of wing veins or the shade of a marking on the thorax, that loss of discriminative detail is exactly where mistakes happen.

Vishwakarma and Ranjan’s answer is an end-to-end architecture built on a modified MaxViT backbone, a vision transformer design first introduced at the European Conference on Computer Vision in 2022. MaxViT is unusual among transformers because it combines two kinds of attention in a single network: window attention, which captures fine-grained local relationships within small patches of the image, and grid attention, which lets distant regions of the image exchange information. This multi-axis attention is what allows the model to notice both the texture of an insect’s body and its overall shape and position in the frame, two kinds of evidence that classical convolutional networks often handle separately.

The key innovation, however, lies in what the authors call a Multi-Stage Feature Fusion Head. Deep networks process images hierarchically: early layers detect simple edges and textures, middle layers assemble those into parts and patterns, and deep layers encode abstract semantic concepts. In many architectures, the low-level texture information is effectively discarded by the time the network reaches its final decision. The fusion head explicitly reintegrates these complementary levels, combining low-level texture features with high-level semantic features so that the final classification draws on the full spectrum of visual evidence. For visually similar pest categories, this fusion proved decisive, since fine surface texture often carries the discriminating signal that abstract features alone cannot supply.

Training such a system on a specialized agricultural dataset presents its own challenge: there is simply no jute pest corpus large enough to train a transformer from scratch. The researchers therefore employed a two-phase transfer learning strategy, starting from representations pretrained on ImageNet, the massive general-purpose image database that has become the standard starting point for computer vision. In the first phase, the pretrained backbone is adapted to the general characteristics of the target domain, stabilizing optimization before the model is asked to master the finer distinctions among pest species. In the second phase, the network fine-tunes on the jute pest data itself, allowing the ImageNet-learned features to reshape around the specific visual grammar of insects on leaves and stems. This staged approach avoids the instability that can occur when a large pretrained model is abruptly exposed to a narrow new domain.

The evaluation was carried out on the publicly available Jute Pest Dataset hosted in the UCI Machine Learning Repository, which covers seventeen jute pest species. The results were striking across every standard metric: overall classification accuracy of 98.43 percent, precision of 98.28 percent, recall of 98.28 percent, and an F1-score of 98.26 percent. The tight agreement among these numbers matters, because it indicates the model performs consistently across all seventeen classes rather than excelling on a few common pests while failing on rare ones. To guard against the possibility that a single favorable train-test split had flattered the results, the authors also ran five-fold stratified cross-validation, and the model’s performance held steady across different data partitions, confirming genuine generalization rather than memorization.

The comparison against established baselines was equally one-sided. The proposed framework consistently outperformed VGG16, ResNet50, DenseNet201, the original Vision Transformer, the Swin Transformer, and ConvNeXt, a lineup that spans a decade of computer vision architecture design. It also surpassed the previously best-performing published method for this task, a feature fusion framework paired with a mountain gazelle optimizer that had reported 96.78 percent accuracy. The new system’s improvement to 98.43 percent represents a meaningful gain in a domain where the remaining errors often involve the most visually confusable species pairs, precisely the cases where misidentification leads to the wrong pesticide or the wrong intervention.

Accuracy alone, however, is not enough for a tool that farmers and agronomists are expected to trust. A model that achieves high scores by latching onto background clutter, such as the color of the soil or the lighting conditions under which photos were taken, will fail unpredictably in the field. To probe what the network was actually looking at, the researchers applied SHAP, or SHapley Additive exPlanations, a technique from game theory that assigns each input region a contribution score for the model’s decision. The analysis showed that the model concentrates its attention on biologically meaningful pest characteristics rather than irrelevant background regions. That transparency is a quiet but important result: it suggests the network has learned something resembling an entomologist’s eye, and it gives practitioners a way to verify that predictions rest on the insect itself.

The broader context is the economics of jute and the promise of precision agriculture. Pest infestations can significantly reduce both crop yield and fiber quality, and jute remains a cornerstone crop for Bangladesh and eastern India, where it sustains farming livelihoods and supplies the global market for natural fiber. Automated, image-based pest identification could allow farmers to detect infestations early, target interventions only where needed, and reduce the blanket pesticide spraying that carries both financial and environmental costs. The authors note that their framework achieves a practical balance between classification accuracy, computational efficiency, and interpretability, the three qualities that determine whether a laboratory model can survive deployment in an intelligent agricultural decision support system.

There are, of course, the usual caveats that separate a benchmark result from a field-ready product. The model was trained and tested on a curated public dataset, and real-world deployment will confront images taken with varied cameras, under harsh sunlight, of partially obscured insects on wind-blown plants. The researchers themselves frame the work as a foundation for practical jute pest monitoring rather than a finished deployment. Still, the trajectory is clear: a single end-to-end transformer, trained with a disciplined two-phase strategy and equipped with a fusion mechanism that preserves both texture and meaning, now classifies seventeen pest species more reliably than any method before it. For a crop that has been cultivated for millennia and identified by eye for just as long, the golden fiber may soon have a digital sentinel watching over it.

Subject of Research: Deep learning-based automated recognition of jute pest species for sustainable agriculture

Article Title: A hierarchical feature fusion transformer for robust Jute Pest recognition in sustainable agriculture

Article References: Vishwakarma, A. K., & Ranjan, R. (2026). A hierarchical feature fusion transformer for robust Jute Pest recognition in sustainable agriculture. International Journal of Machine Learning and Cybernetics, 17(10), Article 494. https://doi.org/10.1007/s13042-026-03330-y

Image Credits: AI Generated

DOI: 10.1007/s13042-026-03330-y

Keywords: jute pests, vision transformer, MaxViT, feature fusion, deep learning, precision agriculture, transfer learning, SHAP explainability, image classification, crop protection, computer vision, sustainable farming

At the heart of the biological argument is a reframing of the soil itself. For most of modern agricultural history, the soil microbiome was treated as passive background. The review positions it instead as a primary determinant of agroecosystem sustainability, with bacteria, fungi, archaea and protozoa driving nutrient cycling, plant growth regulation, disease suppression and resilience to drought and salinity. Through biological nitrogen fixation, phosphorus solubilization and organic matter decomposition, these communities regulate the very nutrients on which rice depends, while also shaping root development, hormonal balance and immune responses. Crucially, the authors stress that microbial abundance alone tells researchers little; the future depends on deciphering microbial function, not merely cataloguing microbial diversity.

One of the most striking mechanisms involves plant growth-promoting rhizobacteria equipped with the enzyme ACC deaminase. When rice plants experience drought, salinity or heavy metal stress, they accumulate ethylene, a stress hormone that in excess stunts root growth and accelerates senescence. ACC deaminase-producing bacteria, including species of Pseudomonas and Bacillus, degrade ACC, the immediate precursor of ethylene, into ammonia and alpha-ketobutyrate. This biochemical intervention keeps roots elongating and taking up water under conditions that would otherwise shut them down. The bacteria also prime antioxidant defences, inducing enzymes such as catalase and superoxide dismutase that neutralize the reactive oxygen species responsible for damaging membranes, proteins and DNA during stress.

The review highlights a decisive shift away from single-strain inoculants, which often perform brilliantly in the laboratory but inconsistently in the field, toward synthetic microbial communities, or SynComs, that combine organisms with complementary metabolic functions. The evidence is compelling. In a greenhouse pot experiment, a SynCom of Azotobacter, Azospirillum and Bradyrhizobium strains derived 51.6 percent of root nitrogen from atmospheric fixation, compared with 25 to 45 percent for individual bacteria. In acidic field conditions, a consortium of aluminium-resistant bacteria raised single-panicle rice yields by 26.36 percent by easing soil acidification and aluminium toxicity. In Vietnam, inoculation with Gluconacetobacter diazotrophicus supported normal growth and yield with just 50 kilograms of nitrogen per hectare, half the conventional rate. Metagenomics now allows researchers to identify the gene markers behind nitrogen fixation, phosphate solubilization and siderophore production, moving the field from descriptive diversity studies toward predictive, function-oriented microbiome engineering.

The biological toolkit extends well beyond fertilization. Microbial biocontrol agents suppress pathogens through antibiosis, nutrient competition, niche exclusion and induced systemic resistance. Disease-suppressive soils rely on specialized metabolites such as 2,4-diacetylphloroglucinol, which disrupts fungal cellular processes; lipopeptides, which compromise pathogen membranes; and hydrocyanic acid, which interferes with respiration. Entomopathogenic fungi like Beauveria bassiana and Metarhizium anisopliae penetrate insect cuticles and kill major rice pests, including the brown planthopper and stem borers, with minimal harm to non-target organisms. Delivery, however, remains the weak link: encapsulation in alginate, chitosan or gum Arabic protects inoculants from desiccation and ultraviolet radiation, and in one rice study, alginate-encapsulated Priestia aryabhattai boosted plant dry mass by 74.9 percent and nitrogen uptake by 86.1 percent compared with free cells.

On the engineering side, the review catalogues a digital transformation that is already reshaping rice paddies. Unmanned aerial vehicles fitted with multispectral and hyperspectral sensors can estimate chlorophyll content, biomass and nitrogen status before deficiencies become visible. In field experiments in Southwest China, UAV remote sensing combined with a random forest model predicted the rice nitrogen nutrition index with R-squared values reaching 0.97, enabling targeted top-dressing recommendations. Normalized Difference Vegetation Index-guided systems direct urea and diammonium phosphate to nutrient-deficient zones, cutting runoff and waste. Digital twin models go further, integrating real-time environmental data with metatranscriptomic information on microbial gene expression, so that irrigation and fertilization decisions account for rhizosphere biology as well as soil moisture.

Water and energy innovations complete the non-microbial picture. Alternate wetting and drying irrigation, in which fields are deliberately allowed to dry between floodings, reduced methane emissions by 47.47 percent in a global meta-analysis spanning 93 percent of world rice production, though nitrous oxide emissions rose by 52.20 percent, and severe drying cut grain yields by 22.6 percent. Mild AWD, by contrast, maintained yields while saving 23.4 percent of water. Solar-powered irrigation systems delivered a higher benefit-cost ratio than diesel pumps in Bangladeshi Boro rice production, while rice husk gasification converts residues that would otherwise be burned into electricity and heat. Biochar emerges as a bridge between the two worlds: its porous structure sequesters carbon, retains nutrients and, critically, acts as a biological scaffold that shelters introduced microbial consortia in the field.

The authors are candid about the bottlenecks. Microbial inoculants still fail unpredictably outdoors because of priority effects, niche exclusion and competition with native communities, and their recruitment patterns shift across plant developmental stages, undermining one-size-fits-all products. Host genotype matters enormously: a formulation tuned to one cultivar’s root exudates may colonize poorly on another, and a line bred for acidic soils may fail in saline paddies. Genome-edited microorganisms promise precision but raise biosafety questions about horizontal gene transfer and face fragmented regulation, with the European Union, the United States and India each applying different oversight frameworks. On the digital side, high equipment costs, weak rural connectivity and limited digital literacy exclude precisely the smallholder farmers who need these tools most; a Malaysian survey found middle- and high-income farmers 20 to 22 percent more willing to pay for IoT technologies than their lower-income peers.

The review’s central message is that integration, not replacement, will define the future of rice. Sensor networks can identify where inoculants are most likely to establish; AI-driven decision-support systems can time applications to coincide with favourable soil moisture and nutrient conditions; biochar and conservation tillage can create habitats that keep engineered communities alive; and CRISPR-assisted breeding can produce rice varieties that actively recruit beneficial microbes through tailored root exudates. Emerging directions include stress-responsive SynComs designed to persist through drought and flooding, methanotrophic bacteria that oxidize methane before it escapes paddies, and circular systems in which straw becomes biochar, biochar carries the microbes, and the microbes feed the next crop. Policy will determine the pace: Vietnam’s One Million Hectares of High-Quality and Low-Emission Rice programme, backed by the International Rice Research Institute, shows how national green-growth strategies can pull these technologies from the laboratory into millions of fields. If the authors are right, the next green revolution will not be written in a single molecule, but in the coordinated choreography of genomes, microbes, satellites and sunlight.

Subject of Research: Integrated microbiome and green technologies for sustainable rice production

Article Title: Integrated Microbiome and Green Technologies for Sustainable Rice Production: Current Advances, Challenges and Future Perspectives

Article References: Aminurrasyid, A. H., Nadarajah, K. K., & Ikmal, A. M. (2026). Integrated Microbiome and Green Technologies for Sustainable Rice Production: Current Advances, Challenges and Future Perspectives. Journal of Agriculture and Food Research, Article 103359. https://doi.org/10.1016/j.jafr.2026.103359

Image Credits: AI Generated

DOI: 10.1016/j.jafr.2026.103359

Keywords: rice, microbiome, SynComs, biofertilizers, precision agriculture, alternate wetting and drying, biochar, CRISPR, methane emissions, solar irrigation, digital agriculture, sustainable farming

The new study, led by a research group at the University of Arkansas and published in the journal Nature, focused on nitrogenase, the enzyme that performs biological nitrogen fixation. Working with a methane-producing microbe known as a methanogen, the collaborative team used high-resolution cryo-electron microscopy to peer into the molecular architecture of the enzyme in unprecedented detail. What they found was unexpected: methanogen nitrogenase can bind with regulatory proteins to form an enormous inactive assembly, which the authors describe as a protein supercomplex. This structure, captured in striking detail by the imaging technique, represents an entirely new way that cells can shut down one of the most energy-demanding reactions in biology.

The supercomplex functions, in essence, as a molecular off switch. When energy or nutrients run low, the nitrogenase enzyme is locked into this inactive state, preventing the microbe from wasting precious resources on a process it cannot afford. Conversely, signals that reflect the cell’s energy and nutrient levels can break the supercomplex apart, freeing the enzyme and allowing nitrogen fixation to resume. This elegant push-pull mechanism gives methanogens a sophisticated level of control over a reaction that consumes large amounts of cellular energy, ensuring that the costly business of splitting atmospheric nitrogen is undertaken only when conditions genuinely permit it.

Dan Lessner, professor of biological sciences at the University of Arkansas and corresponding author of the study, described the significance of the finding in emphatic terms. According to Lessner, the discovery reveals an entirely new strategy for regulating one of the most important biochemical reactions on Earth. That characterization is hardly an exaggeration. Biological nitrogen fixation underpins agriculture, food production and ecosystems across the planet, and the enzyme responsible has fascinated scientists for decades precisely because it accomplishes, at ambient temperature and pressure, what human industry achieves only through enormous inputs of energy.

The agricultural stakes could hardly be higher. As Lessner explained, the biggest limitation on plant growth is often access to sufficient nitrogen, even when sunlight, water and carbon dioxide are abundant. Without adequate nitrogen, plants simply do not grow to their full potential, particularly when it comes to producing the plant material that feeds humans and livestock. Modern agriculture addresses this shortfall through industrial fertilizer production, in which atmospheric nitrogen is chemically converted into ammonia inside high-pressure reactors powered by fossil fuels. The resulting nitrogen-based fertilizer is then spread across fields in the hope that crops will absorb it before it disappears.

That hope is frequently disappointed. A significant portion of applied fertilizer runs off the land into streams and waterways, triggering a cascade of environmental damage. Excess nutrients fuel algal blooms through a process called eutrophication, degrading aquatic habitats and driving losses of biodiversity in affected ecosystems. The environmental cost of synthetic nitrogen is therefore twofold: the fossil-fuel energy consumed in producing ammonia, and the ecological harm caused when that ammonia escapes into the wider environment. Any technology that could reduce either burden would carry enormous economic, social and environmental weight.

This is where nitrogenase enters the picture. As Lessner noted, the enzyme can perform the conversion of nitrogen gas into ammonia biologically, under standard temperature and pressure, without the punishing conditions demanded by industrial reactors. There is intense interest in understanding exactly how the enzyme works and how it catalyzes this reaction, because that knowledge could pave the way toward using nitrogenase as a catalyst for ammonia production in a far more energy-efficient and environmentally friendly manner. The new structural insights into how the enzyme is switched on and off add a critical dimension to that effort, since any engineered application would need to control when and where the reaction takes place.

The genetic dimension of the research opens even more tantalizing possibilities. Because the studies are genetically based, Lessner explained, the genetic information gleaned from these microbes could potentially be transferred into crop plants such as corn. Plants equipped with the machinery to fix their own nitrogen from the atmosphere would, in principle, no longer depend on external fertilizer, alleviating both the financial burden on farmers and the environmental toll of fertilizer runoff. Such a development, Lessner observed, would carry huge economic, social and environmental impacts, transforming one of the most resource-intensive pillars of modern agriculture.

The research itself is a showcase of modern structural biology. Cryo-electron microscopy, the technique at the heart of the study, allows scientists to flash-freeze protein complexes and image them at near-atomic resolution, revealing structures that would be impossible to capture through older methods. It was this approach that exposed the nitrogenase regulatory supercomplex in the methanogen, demonstrating that the enzyme does not operate in isolation but is subject to a physical locking mechanism assembled from multiple proteins. The work was computational and structural in character, and the authors report no competing interests. Lessner’s co-authors included Rajnandani Kashyap, Thomas M. Deere, Ahmed Dhamad, Melissa Chanderban, Monika Tokmina-Lukaszewska, Brian Bothner and Edwin Antony, with Deere, Dhamad and Chanderban affiliated with the University of Arkansas during the course of the work.

Beyond its immediate applications, the discovery resonates with some of the deepest questions in biology. Nitrogen fixation is an ancient capability, and methanogens are among the most evolutionarily ancient microbes known, so understanding how their nitrogenase is regulated sheds light on the physiology of these organisms, the workings of the global nitrogen cycle, and the very evolution of nitrogen fixation itself. Every new detail about how nitrogenase is built, controlled and deployed brings scientists closer to a complete picture of how life learned to unlock the atmosphere. In an era when sustainable food production and climate-conscious industry are urgent global priorities, a molecular off switch hidden inside a methane-producing microbe may prove to be one of the most consequential biological discoveries of the decade, pointing the way toward crops that feed themselves and ammonia factories that borrow their chemistry from the microbial world.

Subject of Research: Regulation of microbial nitrogen fixation through a nitrogenase regulatory protein supercomplex in methanogens

Article Title: Study finds mechanism regulating nitrogen fixation in microbes

Article References: Study finds mechanism regulating nitrogen fixation in microbes. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: nitrogen fixation, nitrogenase, methanogens, cryo-electron microscopy, protein supercomplex, ammonia, fertilizer, agriculture, nitrogen cycle, biotechnology, sustainable farming, University of Arkansas

The authors trace the problem through three generations of biodiversity metrics. Species-based indices describe community composition with precision but offer no mechanism for prediction. Functional-diversity indices, which characterise the distribution of traits such as body size, feeding mode or dispersal ability within a community, still depend on inventories of species that have already been recorded. Habitat-based metrics, the third generation, replace laborious species surveys with habitat classification, dramatically improving scalability, and they now underpin global policy. The Kunming–Montreal Global Biodiversity Framework, adopted at COP15 in December 2022, formalises habitat-based measures in its monitoring targets for 2030. England’s mandatory Biodiversity Net Gain requirement, in force since 2024 under the Environment Act 2021, calculates standardised biodiversity units from habitat distinctiveness and condition. Yet the evidence for these proxies is sobering: studies show the English metric is associated with measurable gains in plant biodiversity but not in birds or butterflies, and it predicts invertebrate biodiversity poorly.

The deeper issue, the authors contend, is that all three metric generations measure patterns rather than the processes that generate them. Even the ambitious European Biodiversity Observation Network roadmap, with its 84 Essential Biodiversity Variables spanning genes to ecosystems, remains observational by design. Such systems can standardise how biodiversity state and change are measured, but they cannot supply the causal prediction needed to evaluate an individual management decision. Correlative models fitted to observed occurrences project associations between recorded species and recorded conditions, with the projected species pool bounded by the region and period used for fitting. They output expected occurrence, not the conditions required for persistence, and they cannot identify which ecological constraint prevents a management response. The distinction the authors draw is not between description and prediction, but between observation-anchored prediction and mechanistic prediction.

The empirical motivation for FTC comes from a stubborn and poorly explained pattern in two decades of agri-environment research: the same intervention produces different, sometimes opposite, biodiversity responses in different landscape contexts. Meta-analyses indicate that organic farming increases species richness by roughly 30 percent on average relative to conventional management, yet 16 percent of studies report negative effects. Benefits tend to be greater in intensively managed, structurally homogeneous landscapes than in heterogeneous ones. A global analysis of 60 crops across six continents confirmed that ecological benefits were greatest in intensive agricultural landscapes, while the same landscape variable could have contrasting effects on different sustainability outcomes. Agri-environment schemes show the same context dependence: a landmark five-country study found substantial variation among countries, taxa and scheme types, with common species benefiting more than species of conservation concern, and a subsequent meta-analysis showed that species richness increased in simple cropland landscapes but not in complex ones.

Spatially explicit modelling offers independent support for a capacity-based interpretation. Across 611 simulated landscapes, landscape structure explained 84 percent of variation in mean beetle density and 70 percent of variation in pesticide impact, with mitigation effects also strongly landscape dependent. The authors argue that these variable outcomes are unlikely to be failures of intervention design. Instead, an intervention acts on biodiversity capacity, a property generated by the landscape as a system, whereas evaluation usually records species occurrence at particular places and times. A predictive metric must represent the processes that generate capacity, not only the patterns those processes have already produced.

So what exactly is Functional Trait Capacity? The authors define it as the range of viable functional-trait combinations that a landscape can sustain, given specified landscape configuration, management regime and assessment horizon. The object of measurement shifts from the community currently observed to the breadth of ecological strategies and life-history solutions for which the landscape provides viable conditions. A wider trait space can support a greater range of species and functional groups, because landscapes act on the resources, disturbance regimes and environmental constraints that filter trait combinations; species composition is one historically contingent outcome of that filtering. The concept builds on Hutchinson’s classic n-dimensional hypervolume and on modern functional-trait ecology, joining them by asking which trait combinations a landscape can sustain rather than which happen to be present.

The framework distinguishes three capacity states. Theoretical FTC is the capacity expected from a site’s physical characteristics and natural ecological trajectory in the absence of deliberate management, a declared reference rather than an absolute biophysical maximum. Management-determined FTC is the capacity produced by a specified land-use and management regime, and it is not necessarily lower than the reference: management can create conditions absent from the natural trajectory, for example by maintaining open habitat where woodland would otherwise develop. Realised biodiversity is the community currently present after historical filtering, the state measured by existing metrics. The gaps between these states represent biodiversity debts or opportunities, separating capacity lost or created through management from capacity that is viable but not yet realised. Crucially, FTC is designed to complement, not replace, observational systems: monitoring tracks realised state and trend, while FTC asks the adjacent causal question of what functional capacity specified conditions should produce.

To show the concept is calculable, the team built a deterministic demonstration using 80 virtual strategies generated from 16 archetypes defined by trophic role, fast–slow life history and disturbance response. Four traits determine viability: spatial access, temporal resource requirement, disturbance tolerance and biotic dependency, with rules for phenological overlap and acyclic food-web structure excluding biologically arbitrary combinations. Four 50-by-50-cell agricultural landscapes varied field configuration and non-crop area in a two-by-two design. The striking result: the researchers selected the same 32-strategy realised community in all four landscapes, yet full assessment of all 80 admissible strategies gave FTC values of 46, 67, 44 and 65. Identical observed richness concealed a difference of up to 23 viable strategies, strategies that occur in none of the observations and therefore could never be recovered by a model fitted only to them. When the same intervention, converting 5 percent of cropped cells to semi-natural habitat along field edges, was applied everywhere, FTC rose by 14 strategies in one landscape, four in another, and not at all in the remaining two, even though mean viable area increased in every case.

Robustness analyses suggest the pattern is not an artefact. Across 60 combinations of persistence thresholds, conversion intensities and placement rules, the high-non-crop landscape always had greater FTC than its low-non-crop counterpart, and intervention effects differed among landscapes in 50 of 60 combinations. In sensitivity ensembles varying the sampled strategies and seasonal resource profiles, baseline FTC differed among landscapes in every replicate, and the intervention response differed among landscapes in all 20 combined replicates. The authors are careful about what this proves: the demonstration shows FTC can be calculated, not that it is empirically valid. Observations remain essential to parameterise resource use, movement and disturbance responses, to test predicted viability and to estimate uncertainty. Their role changes rather than disappears, informing and testing the ecological processes that the model then uses to predict viable strategies under conditions not yet observed.

The path to operational use runs through existing infrastructure. The authors envisage a process-based estimation layer supplying maps, scenario contrasts, limiting-factor diagnoses and uncertainty-bounded indicators to advisers, conservation managers and policy analysts, who would never need to run the underlying model. Inputs correspond to established data streams: Integrated Administration and Control System records for cropping, remote sensing and habitat inventories for configuration, and open trait compilations. Spatially explicit platforms such as the Animal Landscape and Man Simulation System, already applied in eleven European countries, could serve as the mechanistic engine. Practical constraints remain, including restricted access to management records in some member states and incomplete representation of pesticide mixtures and habitat condition. But the conceptual stakes are high. As the EU Nature Restoration Regulation and nature-related financial disclosure frameworks demand choices about future outcomes rather than statements of present state, FTC offers something monitoring alone never can: a falsifiable, scenario-conditioned estimate of what our landscapes could sustain if we managed them differently.

Subject of Research: A computational framework for predicting landscape biodiversity capacity from functional traits and management scenarios

Article Title: Functional Trait Capacity: from observing biodiversity to predicting landscape capacity

Article References: Topping, C. J., Dalgaard, T., Rasmussen, C., & Dupont, Y. L. (2026). Functional Trait Capacity: from observing biodiversity to predicting landscape capacity. npj Sustainable Agriculture, 4(1), Article 79. https://doi.org/10.1038/s44264-026-00190-5

Image Credits: AI Generated

DOI: 10.1038/s44264-026-00190-5

Keywords: biodiversity, functional traits, agriculture, landscape ecology, ecological modelling, agri-environment schemes, conservation policy, Kunming-Montreal Global Biodiversity Framework, sustainable farming, ecosystem management, ALMaSS, biodiversity metrics

The research team, led by Mengxue Zhang and Rong Sheng of the Institute of Subtropical Agriculture at the Chinese Academy of Sciences, built their experiment on a five-year field trial at the Taoyuan Agroecosystem Research Station in Hunan Province. There, plots had received straw at rates corresponding to 0, 50, 100, and 150 percent of the local rice straw yield. To isolate the biological mechanisms from the noise of weather and field variability, the scientists brought soil from each plot into the laboratory and ran controlled microcosm incubations lasting 91 days, with a 28-day flooded period followed by 63 days of drying—mirroring the wet-dry rhythm of a real rice season.

The fertility results were unambiguous. Five years of straw return had left its signature on the soil: bulk density fell, while cation exchange capacity, total carbon, total nitrogen, available nitrogen, available phosphorus, and available potassium all climbed with increasing straw rates. Even the lowest input, the half-rate treatment, cut bulk density by nearly 3 percent and lifted available potassium by more than 38 percent, available phosphorus by about 15 percent, and available nitrogen by nearly 7 percent relative to the no-straw control. Micronutrients such as manganese, copper, and zinc also rose. In short, the soil was demonstrably better at holding water, nutrients, and cations—the hallmarks of fertility.

But the greenhouse gas picture told a sharply different story depending on dose. Carbon dioxide fluxes rose in proportion to the straw input, with the half-rate treatment emitting 66 percent more CO2 than the control and the full and 150-percent rates reaching 2.2 and 2.5 times the control level. Methane, the more potent concern, was even more dose-sensitive: cumulative emissions from the half-rate treatment stayed low at 4.41 grams per square meter, while the full-rate and 150-percent treatments belched out roughly 23 and 38 grams per square meter—four and eight times higher, respectively. When the researchers converted all three gases into a single global warming potential metric using the latest 100-year conversion factors, the half-rate treatment showed only a 3 percent, statistically insignificant increase over the control, whereas the higher rates inflated the warming potential by up to 151 percent.

The key to this asymmetry lay in nitrous oxide, a greenhouse gas nearly 273 times more powerful than CO2 over a century. Counterintuitively, the half-rate treatment emitted significantly less N2O than the unfertilized-with-straw control—just 1.63 grams per square meter. The explanation emerged from the microbial genes the team quantified. N2O is produced by denitrifiers carrying the nirS and nirK genes and consumed by microbes carrying nosZ-type genes encoding N2O reductase. The half-rate soil harbored the highest abundance of the nosZII clade and the lowest nirS-to-nosZII ratio of any treatment, around 0.81 compared with 1.11 to 1.36 elsewhere. That combination—modest production capacity paired with maximal consumption capacity—meant the soil was effectively eating its own N2O before it could escape.

The timing of emissions added further nuance. Two distinct N2O peaks appeared during the incubation: one about 11 days after fertilization under flooded conditions, and another during the late drying phase. At the first peak, the control plot emitted the most N2O, likely because urea nitrogen sat available for nitrifiers and denitrifiers, while in straw-amended soils proliferating microbes competed fiercely for that same nitrogen. At the second peak, during drying, the order reversed as oxygen penetrated the profile and organic matter mineralization released fresh nitrogen. Yet even here, the half-rate soil kept its N2O fluxes as low as the control’s, thanks to its superior N2O-reducing community.

Methane told its own mechanistic tale. Methanogens, flagged by the mcrA gene, multiplied in proportion to straw input, with average abundances during the flux period rising 32 percent at the half rate and 129 percent at the highest rate. But production alone does not determine emissions—consumption by methanotrophs carrying the pmoA gene matters just as much. The researchers identified a critical threshold: a soil methane holding capacity of roughly 190 millimoles per mole under anaerobic conditions. Below that ceiling, retained methane could be oxidized in place; above it, the excess escaped to the atmosphere. The half-rate treatment’s soil methane concentrations fluctuated right around 200 millimoles per mole, hovering at the edge, while the higher rates blew well past the capacity. Meanwhile, the half-rate soil boasted methanotroph abundances about 81 percent higher than the control, giving it the oxidative muscle to mop up most of what its methanogens produced.

The broader significance of these findings is considerable. Rice cultivation accounts for roughly 12 percent of global agricultural greenhouse gas emissions, and straw incorporation has been identified as one of the key drivers, with reported methane increases in prior studies ranging from 23 percent to a staggering 1,192 percent. By demonstrating that the dose of straw—not merely its presence—governs the climate outcome, the study offers farmers and policymakers a concrete, actionable target: return about half the harvest. Previous work had hinted at this, showing that half-rate incorporation cut methane emissions by 48 percent relative to full incorporation, but the new research goes further by tying the climate benefit to specific microbial mechanisms and by simultaneously confirming that fertility gains are preserved.

The study also resolves a long-standing controversy in the literature. Reports on how straw affects N2O emissions have swung wildly, from 73 percent reductions to 137 percent increases, and the abundances of N2O-related genes have been reported to rise, fall, or remain unchanged. The Hunan results suggest that much of this inconsistency may stem from differing straw application rates across studies, which shift the balance between N2O-producing nirS communities and N2O-consuming nosZII communities in opposite directions. Similarly, conflicting findings on methanotroph responses—some studies found no change in pmoA abundance, others increases of up to 59 percent—may reflect the same dose dependence.

The authors caution that their conclusions come from a single soil type, an acidic Ultisol derived from quaternary red clay under a subtropical monsoon climate. Different soils possess different methane holding capacities and nitrogen transformation dynamics, so field tests across diverse soil types remain necessary before the half-rate prescription can be generalized worldwide. Still, the core message stands as a rare piece of good news in agricultural climate science: with the right amount of straw, rice paddies can grow more fertile without growing warmer, turning an agricultural waste problem into a climate-smart resource.

Subject of Research: Optimal rice straw incorporation rates for balancing paddy soil fertility and greenhouse gas emissions

Article Title: Mechanisms of enhancing soil fertility without obviously elevating global warming potential under an optimal rice straw incorporation rate in a paddy soil

Article References: Zhang, M., Liao, R., Zhang, W., Fang, C., Guerrero-Cruz, S., Táncsics, A., Zhu, B., Wei, W., & Sheng, R. (2026). Mechanisms of enhancing soil fertility without obviously elevating global warming potential under an optimal rice straw incorporation rate in a paddy soil. SOIL, 12(2), 871-883. https://doi.org/10.5194/soil-12-871-2026

Image Credits: AI Generated

DOI: 10.5194/soil-12-871-2026

Keywords: rice straw, paddy soil, soil fertility, greenhouse gases, methane, nitrous oxide, methanotrophs, denitrification, global warming potential, climate-smart agriculture, soil microbiology, sustainable farming