The expanding global population and intensified land use are placing agroforestry systems under growing strain (Blaser et al., 2018). According to projections by Tilman et al. (Tilman et al., 2011), global crop demand is set to double between 2005 and 2050. Satisfying this surge in demand within the planet's finite spatial and temporal boundaries represents one of the century's most pressing challenges. Within this context, the development of robust plant disease management protocols has gained critical urgency (He et al., 2016). Indeed, pathogens, as the primary agents of plant disease, cause severe declines in both crop yield and quality, directly threatening food security and agricultural economies worldwide (Zarco-Tejada et al., 2021). Savary et al. estimate that diseases are responsible for roughly 40 % of global crop losses annually, translating to economic damages in excess of $220 billion (Savary et al., 2019). This diverse group of pathogenic microorganisms, which includes fungi, bacteria, oomycetes, viruses, and nematodes, invades plant tissues to disrupt normal physiological functions, thereby inducing a spectrum of disease symptoms (Lamichhane et al., 2024). Compounding the problem, frequent outbreaks often drive farmers to increase their dependence on broad-spectrum fungicides. However, this response risks inflicting long-term, and sometimes irreversible, harm on surrounding ecosystems (Zhang et al., 2024a). Hence, curbing disease progression promptly and preventing widespread epidemics is essential for securing crop production, preserving ecological stability, and protecting human health.

Traditionally, disease diagnosis in agriculture has relied on farmers' visual inspection of plant parts such as roots, stems, leaves, flowers, and fruits (Flores et al., 2021). Yet, this empirical approach is not only susceptible to human error but also demands significant time and labor (Mahlein et al., 2018). A major complication is that clear visual symptoms typically only appear in the middle to late stages of the disease. By that point, the pathogen may have spread extensively, making it uncontrollable even with heavy pesticide application (Kashyap et al., 2017). Therefore, many growers choose to apply pesticides preventively during the early stages of crop growth, or before any symptoms appear, to mitigate potential disease risks. However, such preventive measures carry significant risks. They may trigger phytotoxic reactions, reduce soil fertility, and lead to long-term consequences including the evolution of resistant pathogen strains and the accumulation of pesticide residues. Integrating the principle of precision management with technologies that enable accurate early-stage detection thus emerges as a critical strategy for advancing sustainable crop production, safeguarding human health, and protecting the environment. At present, precision agriculture supported by intelligent technology is developing rapidly. Against this backdrop, the development of advanced early disease monitoring and warning tools not only fully aligns with the development direction of modern agriculture but is also highly consistent with the broader sustainable development goals.

Research on plant disease management has gone through more than a hundred years. From the late 18th century to the mid-20th century, the diagnosis of pathogens mainly relied on naked-eye observation of symptoms, combined with basic isolation or culture techniques (Zhang et al., 2025a). Although these methods are simple and easy to implement, they take a long time and are difficult to be used for the early identification of diseases (Ray et al., 2017). By the 1970s and 1980s, the emergence of molecular biology techniques brought about a turning point. Immunological detection methods (such as enzyme-linked immunosorbent assay, ELISA) and DNA-based technologies (such as polymerase chain reaction, PCR) opened new avenues for the early and rapid identification of pathogens (Khater et al., 2017; López et al., 2003). Immunoassays employ different classes of antibodies, including monoclonal, polyclonal and heavy-chain types, all of which can recognize specific pathogens with a high degree of selectivity. Nucleic-acid based detection techniques also provide strong specificity, and they offer the additional advantage of lowering the cost of testing materials (Zhang et al., 2025b; Rezavand et al., 2016; Farooq et al., 2018). Meanwhile, the integrated development of biological science, information science and materials science has also greatly promoted the progress of pathogen biosensing technology (Wang et al., 2024a). Biosensors are typically composed of biometric elements (such as antibodies (Berto et al., 2019), aptamers (Hong et al., 2021), enzymes (Zhang et al., 2025a), antimicrobial peptides (Hoyos-Nogués et al., 2018), phages (Samson et al., 2024) or DNA probes (Vaseghi et al., 2013)) combined with electrochemical or optical signal converters, capable of achieving precise and sensitive detection of pathogens at extremely low concentrations. It is worth mentioning that biosensors have successfully broken through the limitation of traditional methods that require “remote submission for testing”, thereby significantly enhancing the timeliness of detection, simplifying the workflow, and reducing the reliance on expensive equipment and professional operators. The latest research in plant pathology has further revealed multiple response mechanisms of plants in response to pathogen invasion. Similar to animals, plants also undergo a series of subtle physiological and phenotypic changes when they are attacked (Dodds et al., 2024; Ye et al., 2021), such as the release of specific signal molecules, characteristic acoustic wave vibrations in leaves or fruits, and minor changes in appearance traits. Therefore, in the future, if we can deeply understand the biological characteristics of pathogens (such as their types, infection sites, transmission methods, and molecular interaction mechanisms with plants), and develop the ability to remotely interpret plant stress signals, it is expected to effectively bridge the technological gap between precise laboratory diagnosis and large-scale real-time field monitoring.

Numerous review articles have explored plant disease detection technologies and their application potential. For instance, Zhou et al. summarized research progress on monitoring small molecule changes within plants using wearable electrochemical sensors to predict disease status (Zhou et al., 2024a). Giraldo and colleagues examined how advances in smart nanobiotechnology can be combined with electronic platforms to monitor molecular signals released by plants experiencing biotic stress (Giraldo et al., 2019). In a related study, Steeneken and co-authors discussed the promising role of wireless sensor networks in agriculture and stressed the need for an accurate understanding of how these sensing tools should be applied in practice (Steeneken et al., 2023). Despite these contributions, most existing reviews concentrate on a single technological direction or sensing strategy and do not fully capture the underlying links between pathogen biology and sensor design. Few studies also provide a balanced assessment of how feasible different sensing approaches are when they are deployed in field settings. In addition, the large number of plant diseases, the complexity of their infection processes and the variety of sensing technologies available for early detection create substantial difficulties for researchers and practitioners.

Hence, this review seeks to clarify recent progress in sensor technologies that support the detection of plant pathogen infections. We begin by organizing the major categories of plant pathogens and summarizing their biological infection characteristics along with process-related signaling molecules. This overview also considers how a deeper understanding of these processes can guide the development of new sensing approaches. In the next part, we group pathogen sensors into two broad categories according to their detection targets and underlying mechanisms. The first group involves direct detection that focuses on the pathogens themselves, and the second group involves indirect detection that captures plant responses to infection as shown in Fig. 1. For each category, we discuss the operating principles, key strengths and practical constraints, and we evaluate their potential use in future agricultural settings. The review concludes with a summary and perspective on possible research directions. The goal is to assist scientists and agricultural professionals in comparing available sensing technologies, selecting suitable options for their needs and improving the accuracy and efficiency of plant disease monitoring and management.

The contributions of this review can be summarized in three main aspects. First, at the conceptual level, we organize recent advances within a “pathogenesis-to-sensor” framework that explicitly links pathogen biology to sensor design. We discuss how infection routes, effector molecules and pathogen-derived signals influence the selection of sensing materials and device architectures, so that sensor development can be more firmly grounded in the mechanisms of disease progression. Second, in terms of scope, we bring together both pathogen-targeted biosensors and plant-based indirect sensing approaches, and assess their performance in realistic agricultural scenarios. The discussion covers optical and electrochemical platforms, VOC-based sensing, acoustic sensing, remote sensing and other emerging techniques, with attention to cost, detection performance, ease of operation and stability in the field. Third, from an application perspective, we outline a simple decision-making framework that follows the sequence of screening, analysis and verification. This framework aims to guide users in matching sensing strategies to practical needs, and to build a closed-loop monitoring pathway from regional screening to precise on-site diagnosis for early warning and control of plant diseases.