A research team led by Ruifeng Guo and Yonglin Wang of Beijing Forestry University, together with colleagues at the Beijing Academy of Agriculture and Forestry Sciences and the United States Department of Agriculture, set out to map the microbial communities living inside smoke tree branches. The researchers sampled twelve trees, six healthy and six diseased, separating each branch into two distinct compartments: the outer epidermis, or bark, and the inner xylem, the vascular tissue that transports water. From each tree they collected branches in three health states: branches from completely healthy trees, symptomless branches from diseased trees, and the visibly diseased branches themselves. Using amplicon sequencing of bacterial 16S rRNA and fungal ITS regions, they generated a detailed census of the endophytic microbiota inhabiting each tissue type.

The sequencing data revealed that Verticillium wilt profoundly reshapes the internal microbial world of the tree. Bacterial communities in diseased xylem actually showed higher Shannon diversity than in healthy xylem, suggesting that infected trees may recruit additional microorganisms as a defensive response. Fungal communities told the opposite story: diversity dropped significantly in the epidermis and xylem of diseased branches. Principal coordinate analysis and permutational multivariate analysis of variance confirmed that disease status significantly altered community composition in nearly every compartment. Critically, when the researchers re-ran their analyses after excluding V. dahliae sequences, the pattern held, meaning the pathogen’s influence on the fungal community extended well beyond its own abundance. Quantitatively, disease explained roughly 40 percent of the variation in fungal communities but only about 11 percent in bacterial communities, indicating that fungi are far more sensitive to the wilt epidemic than bacteria.

To probe the stability of these communities, the team constructed co-occurrence networks linking microbial taxa based on correlated abundances. The fungal networks in diseased branches had fewer connections, fewer negative correlations, and markedly lower robustness when nodes were removed in simulations of species extinction. Both diseased and symptomless branches from infected trees showed elevated vulnerability, meaning the loss of a single taxon could disproportionately collapse network efficiency. Fungal networks proved consistently more fragile than bacterial ones, echoing earlier findings in pepper stems and olive roots where disease also destabilized fungal assemblages. This loss of connectivity, a form of microbial dysbiosis, suggested that wilt disease does not merely add a pathogen to an otherwise intact community; it actively dismantles the ecological architecture that keeps endophytes in check.

Within this destabilized landscape, one genus stood out. Differential abundance analysis and random forest classification both identified Botryosphaeria as a top biomarker of diseased branches, enriched alongside Verticillium itself. Linear correlation analysis showed that the relative abundance of Verticillium was positively and significantly associated with Botryosphaeria, with an R-squared of 0.47, while it was negatively correlated with Alternaria and Cladosporium. The team isolated a strain of Botryosphaeria dothidea from infected branches and found that it showed no antagonistic activity against V. dahliae in culture, ruling out simple competition. Spatially, the two fungi occupied different niches: B. dothidea concentrated in the bark while V. dahliae dominated the xylem. Field collections from four Beijing districts reinforced the association, with both fungi recovered together from 37 percent of 76 diseased branch samples and V. dahliae alone from 63 percent.

The decisive test came in the greenhouse. When smoke tree seedlings were inoculated with B. dothidea alone, nothing happened; the trees stayed healthy. When they were inoculated with V. dahliae alone, characteristic wilt symptoms appeared. But when seedlings received V. dahliae first and B. dothidea second, wilting was dramatically more severe than with either pathogen alone. Quantitative PCR revealed the mechanism behind this synergy: V. dahliae biomass was unchanged by the presence of its partner, but B. dothidea biomass surged in co-inoculated plants. In other words, the vascular pathogen was not simply teaming up with an equal partner; it was unlocking the door for a latent endophyte to multiply and inflict damage it could never achieve on its own.

Physiological and histochemical assays traced how the trees lost their ability to resist the second invader. Catalase and peroxidase, the antioxidant enzymes that scavenge reactive oxygen species during immune responses, were most active in plants challenged with V. dahliae alone and lowest in co-inoculated plants. Malondialdehyde content and relative conductivity, both markers of cell membrane damage, peaked in co-inoculated seedlings. DAB staining showed that V. dahliae infection drove substantial hydrogen peroxide accumulation, while trypan blue staining revealed extensive cell death, and microscopy documented solid blockage of xylem vessels. Together these results indicated that the first infection had already crippled the tree’s defense system before the second fungus even arrived.

Transcriptome sequencing provided the molecular explanation. In V. dahliae-infected branches, 2,800 predicted genes were significantly downregulated and 1,871 were upregulated compared with controls. Among the downregulated set, 394 genes in a single expression cluster were tied to plant defense, including disease resistance genes, protein kinases, jasmonic acid-responsive factors, chitinase-binding proteins, terpenoid synthases, cellulose synthases, protease inhibitors, redox regulators, and transcription factors. KEGG enrichment showed that hormone signal transduction, phenylpropanoid biosynthesis, MAPK signaling, brassinosteroid and monoterpenoid biosynthesis, and benzoxazinoid production were all significantly suppressed. Gene ontology analysis likewise flagged reduced transmembrane receptor kinase activity and compromised cell wall and cytoskeleton components. B. dothidea-infected branches showed a strikingly similar transcriptional signature, with the same defense and hormone pathways downregulated, confirming that both fungi exploit overlapping vulnerabilities in the host.

The study reframes B. dothidea, long recognized as a worldwide latent pathogen of woody plants that waits for drought, wounds, or stress before striking, as an opportunistic accomplice whose pathogenicity is triggered by another microbe rather than by abiotic stress alone. It also adds smoke tree wilt to a growing list of plant diseases, from grafted grapevine decline to soybean root rot, in which co-infection produces symptoms far worse than the sum of the parts. For forest managers, the practical implications are significant: because the two fungi occupy different tissues, control strategies may need to pair fungicides with measures that boost overall host vigor, keeping the latent bark dweller in its harmless endophytic state. The researchers are already testing plant immune activators in smoke tree plots to see whether shoring up the tree’s own defenses can break the deadly synergy, offering hope that the red hills of Beijing can be protected not by attacking one pathogen, but by managing the entire microbial community within the tree.

Subject of Research: Synergistic co-infection of smoke trees by Verticillium dahliae and the latent bark fungus Botryosphaeria dothidea

Article Title: The bark latent fungus Botryosphaeria dothidea exacerbates branch dieback following the infection with Verticillium dahliae

Article References: Guo, R., Li, Y., Tang, C., Zhao, Y., Wang, M., Qiao, G., Klosterman, S. J., & Wang, Y. (2026). The bark latent fungus Botryosphaeria dothidea exacerbates branch dieback following the infection with Verticillium dahliae. Stress Biology, 6(1), Article 13. https://doi.org/10.1007/s44154-026-00288-3

Image Credits: AI Generated

DOI: 10.1007/s44154-026-00288-3

Keywords: Verticillium dahliae, Botryosphaeria dothidea, smoke tree, Cotinus coggygria, Verticillium wilt, plant microbiome, endophytes, co-infection, branch dieback, xylem, RNA-seq, forest pathology

The platform, called AgroYachay—’yachay’ meaning knowledge in Quechua—is described in the journal SoftwareX and released under the MIT licence with an archived DOI, version 1.0.1. It is the cloud counterpart of a companion tool, AgroCommish, which handles the upstream work of commissioning ESP32 sensor nodes: flashing firmware, discovering sensor pins, provisioning WiFi credentials and registering each verified device with the cloud. Together, the two tools form a complete, reproducible device-to-decision pipeline that a small research group or agricultural programme can deploy entirely on open-source software and commodity hardware, with no per-seat licensing or proprietary cloud lock-in.

Architecturally, AgroYachay follows a three-tier edge-cloud design. At the edge, ESP32 microcontrollers wired to a DHT11 sensor for air temperature and humidity and an FC-28 probe for soil moisture post JSON readings over WiFi every ten seconds. The backend is a Flask 3 application of roughly 7,300 lines of Python, organised into controllers, feature modules and a service layer, with state persisted across fifteen normalised tables in PostgreSQL 17. Authentication relies on JSON Web Tokens with bcrypt password hashing and optional Google OAuth. The frontend is a React 18 single-page application of about 13,000 lines styled with Tailwind CSS, charting live sensor series with Recharts. More than ninety REST API routes tie the system together.

The intelligence layer is deliberately hybrid. Image-based pest diagnosis runs on a self-hosted open multimodal model—qwen2.5-VL-3B, served through Ollama—which keeps farmers’ photographs on local hardware rather than shipping them to a third-party image service. A grower photographs an affected leaf; the vision model, prompted as a phytopathologist specialising in Andean crops, returns a structured verdict covering health status, disease name, confidence, severity, causes, treatment and prevention. A follow-up text-only call to a cloud large language model hosted by Groq converts that verdict into an actionable management plan: urgency, step-by-step actions, chemical products with doses available in Peru, an organic alternative, and guidance on when to consult a human agronomist. The prompt targets potato, quinoa, maize, faba bean, oca and cañihua, the staple crops of the region.

The team validated the vision module with unusual candour, evaluating it as a plant-health triage aid rather than a fine-grained diagnostic tool. Using 509 images—300 laboratory photographs from PlantVillage and 209 field images from PlantDoc—they found perfect sensitivity: every one of the 409 diseased leaves was flagged as diseased, the property that matters most for an early-warning system. Specificity was lower at 62 percent, meaning the model over-flags healthy leaves as diseased, a conservative bias favouring missed nothing over missed disease. Species-level discrimination was weaker, with late blight frequently confused with early blight and a macro-F1 of 0.55. The authors state plainly that the module supports binary field triage, not reliable differential diagnosis, and that performance on real Andean field imagery remains untested. Repeated queries at the deployed temperature of zero produced identical classifications in ten out of ten trials, confirming reproducibility.

Beyond diagnosis, a context-aware conversational assistant answers free-form questions in Spanish, conditioned on the farmer’s registered crops, region and latest sensor values. Separate services translate the five-day OpenWeather forecast into a risk level, a weekly activity plan and optimal-day recommendations for spraying, sowing and irrigation. A crop management module records species, variety, parcel area, GPS location, sowing date and phenological stage, ensuring that every reading, alert, diagnosis and estimate is tied to a specific crop and device. An inputs module calculates fertiliser and agrochemical requirements from parcel area and target yield, while an advisory module queues requests for human agronomists, keeping an expert in the loop when the automated tools should not act alone.

The platform’s most distinctive feature may be its economic layer. Rather than a black-box regressor, the yield-and-revenue estimator is a transparent multiplicative factor model requiring no training data, which suits the sparse-data reality of Andean deployment. A crop-specific base yield—15.5 tonnes per hectare for potato, 1.8 for quinoa, 8.5 for oca—is adjusted by climate factors computed from mean temperature, humidity and accumulated rainfall within phenological windows, by a phenological factor that grows with crop progress, and by a piecewise area factor. The projected tonnage is multiplied by a regional reference price to give expected gross revenue. Because the model is purely multiplicative, its sensitivity is fully transparent: a 10 percent error in any climate factor moves the estimate by exactly 10 percent, and every contributing factor is displayed to the user. Results include pessimistic, probable and optimistic scenarios with a confidence score, and can be exported as styled PDF reports suitable for credit or crop-insurance applications.

The multilingual interface is treated as a first-class feature rather than an afterthought. Of 1,488 UI strings, 94 percent are localised into both Quechua and Aymara, with a complete English and Spanish locale; only technical terms such as ‘IoT’ and ‘ESP32’ are deliberately left untranslated. This matters because a substantial share of Andean smallholders operates primarily in indigenous languages, and generic farm-management platforms assume Spanish or English literacy. The authors note an honest limitation here: while the interface is multilingual, the LLM-generated advice, pest verdicts and exported reports are currently produced in Spanish only, so a Quechua-speaking user still encounters Spanish text at the moment of advice. Extending the prompts and report templates to indigenous languages is the next inclusion step.

Performance measurements on the production server show that core operations are genuinely interactive: lightweight REST calls complete in 2.6 milliseconds at the median, dashboard reads in under 10 milliseconds, and sensor ingestion in 19 milliseconds. A 22-minute bench test of a real ESP32 node delivered 132 of 134 expected sampling cycles, a 98.5 percent success rate with a maximum reconnection gap of 20 seconds. The exception is the vision module: on the CPU-only demonstration server a single diagnosis takes minutes, whereas on a consumer GPU it completes in seconds, so GPU-backed deployment is recommended for interactive use. The conversational assistant, dominated by external inference, averages around two seconds per response.

The developers, based at the Universidad Nacional del Altiplano in Puno, drew requirements from direct field experience in a smallholder economy centred on potato, quinoa, cañihua and Andean livestock, where commercial precision-agriculture suites are cost-prohibitive. They self-tested the platform on their own plantings and received positive informal feedback from local farmer demonstrations, though a formal quantitative user-acceptance study remains future work. So do calibration of the yield estimator against actual harvest records, an expert-labelled Andean field image benchmark, an on-device offline fallback for pest screening, and security hardening measures such as per-device authentication, database encryption at rest and automated secret rotation. Even without the LLM layer, sensor monitoring, alerts, the factor-model estimator and reporting already function. As an early but complete contribution, AgroYachay demonstrates that the entire chain—from flashing a sensor in a highland workshop to issuing a financed-harvest report—can now run on open-source software, potentially reshaping who gets to participate in precision agriculture across Latin America and other developing regions.

Subject of Research: An open-source IoT and large-language-model platform supporting agronomic and economic decision-making for Andean smallholder farmers.

Article Title: AgroYachay: An open-source IoT and large-language-model platform supporting agronomic and economic decision-making for Andean smallholders

Article References: Torres-Cruz, F., Vilca-Solorzano, R. A., Yana-Yucra, D. M., Ibañez-Quispe, V., & Fuentes-Navarro, E. L. (2026). AgroYachay: An open-source IoT and large-language-model platform supporting agronomic and economic decision-making for Andean smallholders. SoftwareX, 36, Article 103037. https://doi.org/10.1016/j.softx.2026.103037

Image Credits: AI Generated

DOI: 10.1016/j.softx.2026.103037

Keywords: AgroYachay, precision agriculture, IoT, ESP32, large language models, smallholder farming, Andean agriculture, plant disease diagnosis, open-source software, yield prediction, Quechua, Peru

The challenge is not simply to determine whether a plant is healthy or diseased. In precision agriculture, an autonomous robot or drone may need to identify the specific disease, estimate the proportion of tissue affected and determine whether symptoms are concentrated near the bottom of the plant or distributed in another spatial pattern. Those details can influence when and where fungicides are applied, potentially reducing chemical use while protecting yields. Yet manual scouting is slow, expensive and dependent on trained specialists. The problem is becoming more urgent as climate change, population growth, shifting agricultural conditions and pathogen evolution create opportunities for diseases to emerge in new regions or appear in unfamiliar forms.

Many existing plant-disease models are based on convolutional neural networks or vision transformers. These systems can perform impressively when trained on large collections of carefully labeled images, but their apparent success often depends on data that do not resemble real farms. Public datasets frequently contain isolated leaves photographed against simple backgrounds, with consistent lighting and clear disease symptoms. Field images, by contrast, may include overlapping foliage, soil, weeds, shadows, glare, changing weather and partially obscured lesions. Building a sufficiently large, diverse and precisely annotated field dataset is costly. It is especially difficult to label subtle severity levels or describe how lesions are distributed throughout an entire plant, since such judgments can be subjective and require considerable expert time.

PD-CLIP addresses one of the central weaknesses of conventional classifiers by using a vision-language architecture inspired by CLIP, or contrastive language-image pre-training. Rather than learning only a fixed list of numerical class labels, the model maps images and textual descriptions into a shared mathematical feature space. During training, image representations are pulled closer to the descriptions that match them, while mismatched image-text pairs are pushed apart. Once this alignment has been learned, the system can compare a new plant image with candidate text prompts and select the description whose representation is most similar. In principle, this open-vocabulary design allows the model to reason about disease categories or traits that were not directly represented in its visual training examples.

The researchers aim to make those text descriptions substantially more informative than generic prompts such as “a diseased leaf.” Their system uses a multimodal large language model to generate structured descriptions containing disease type, severity and spread type. These descriptions are paired with synthetic images generated in Unreal Engine 5, where disease characteristics can be controlled systematically. Three-dimensional physical simulation allows the researchers to construct complete virtual plants, alter the amount of diseased tissue and vary where symptoms appear across different height levels. An iterative texture-overlay process can then place disease-like patterns on plant surfaces, producing whole-plant images and close-up views from the same virtual scene.

This two-scale structure is crucial because plant diagnosis requires both a broad view and a microscopic one. A close image of a leaf may reveal the color, shape and texture of lesions needed to distinguish among visually similar diseases. However, that crop may provide little information about whether symptoms are concentrated at the base of the plant or spread throughout the canopy. A distant image preserves this spatial context but can make small pathological details difficult to see. PD-CLIP therefore processes global images showing the entire canopy alongside local patches that emphasize leaf-level symptoms. The paired observations are intended to connect local pathology with the larger pattern of disease progression.

The framework also incorporates real-world images collected in tomato fields in North Carolina. The field experiments included several tomato varieties and were conducted between June 18 and August 20, 2025, with observations taken at four stages as disease symptoms developed over time. A customized phenotyping platform based on the Amiga robot moved through crop rows while stereo cameras and active strobe lighting captured canopy images from both sides. The active illumination was used to reduce variation caused by changing outdoor light, a common source of failure when a model trained under one set of conditions is deployed under another. In total, the researchers collected 12,096 real-world images covering early blight, late blight and septoria leaf spot, with symptoms categorized across six affected severity levels and a bottom spread type.

Synthetic images and field photographs, however, do not naturally look alike. Virtual plants may have different textures, lighting, backgrounds and distributions of symptoms from those found in real agriculture. This discrepancy is known as a domain gap, and it can cause a model to perform well in simulation but poorly in the field. PD-CLIP uses a composite domain-adaptation strategy to reduce that gap. Such strategies can align image features between source and target domains, reduce low-level differences in appearance, encourage stable predictions on unlabeled target images and refine high-confidence pseudo-labels. They must also preserve class boundaries so that making synthetic and real images more similar does not erase the distinctions between diseases or severity categories.

In the proposed workflow, dual-stream encoders independently process visual information and disease-language descriptions before placing both in a shared latent space through contrastive optimization. Domain adaptation operates alongside this alignment, transferring information learned from controllable synthetic data toward unconstrained field observations. The final system performs zero-shot inference by comparing a new image embedding with candidate text embeddings, without requiring task-specific retraining for every new diagnostic question. The framework is intended to output several traits at once: the disease identity, its severity level and its spatial spread pattern. This differs from many agricultural models that focus on a single leaf, a single disease label or a closed set of categories defined before deployment.

The researchers position PD-CLIP as a data-efficient foundation for agricultural diagnosis rather than a replacement for field experts or a universally solved system. Its significance lies in linking four capabilities that are usually studied separately: controllable three-dimensional data generation, detailed semantic descriptions, efficient vision-language adaptation and transfer from simulation to reality. If validated across broader crops, environments and pathogen classes, such systems could help robots and drones perform more targeted monitoring, identify unusual threats earlier and support variable-rate treatment decisions. The study’s framework does not eliminate the difficulty of real-world disease diagnosis, but it offers a technically ambitious strategy for giving agricultural AI a richer understanding of what symptoms look like, where they occur and what they may mean.