1 Introduction

Feeding the growing human population in a sustainable way and in a changing climate is one of the most important challenges we face in the 21st century. Without substantial shifts in agricultural practices on a global scale, we risk surpassing the planetary boundaries, including freshwater use and biogeochemical flows of nitrogen (N) and phosphorus (P) (Campbell et al. 2017). Currently, the best croplands are already under cultivation, and much of the remaining land is unsuitable for farming (Schneider et al. 2022). In addition, most of this cropland is rain fed, making crops vulnerable to unpredictable weather patterns. Other key inputs controlling plant productivity, such as synthetic fertilizers, are often overused, with only a fraction effectively reaching the plants, while the remainder is lost to the environment (Omara et al. 2019). Moreover, the rate of productivity improvement for many of our most important crops is slowing down (Bowman and Zilberman 2013). Ensuring a healthy future for the people and the planet requires a deeper understanding of crop-based agriculture systems and their response to the environment.

Agriculture involves managing complex living systems embedded within complex ecological, societal, and economic systems that operate at multiple scales (Figure 1). At the farm level, management practices rely on predictive modeling, supported by measurements taken from soils and individual plants and the associated organisms that make up their microbiomes (Shaikh et al. 2022). Transitioning to a regional scale, effective coordination of shared resources such as land and water, alongside the development of policies, is essential to ensure sustainable production. On the global scale, concentrated efforts are necessary to address challenges including climate change and equitable food distribution, thereby ensuring food security worldwide. Changing agricultural practices requires intense collaboration of all disciplines involved in these systems, including plant science, engineering, and computer science. Furthermore, given that agricultural outcomes impact everyone, this transdisciplinary effort must involve diverse fields of economics, policy, social sciences, and the engagement of various stakeholders including farmers and consumers.

The complexity of agricultural systems, including plants, their interaction with a local environment that spans from the atmosphere to below ground, and the human context in both management and consumption, presents numerous scientific and technical challenges in developing sustainable agricultural practices. The widespread use of synthetic fertilizers, primarily containing nitrogen and phosphorus, for decades has significantly boosted global food production capacity (Steffen et al. 2015). However, a major challenge arises from the inefficient application of fertilizers, which contaminates the environment via leaching, particularly in water bodies where it damages aquatic ecosystems (Schindler 1974). For example, crops utilize only around one-third of the nitrogen supplied through fertilization, with the remainder lost through leaching and gaseous emissions. Phosphorus fertilization has also exceeded crop phosphorus use, with only 10%–20% of phosphorus fertilizer being utilized by crops (Bovill et al. 2013). As we lack tools capable of providing real-time information on the internal state of plants and fertilizer availability in the soil, we are unable to develop predictive models to guide farmers in effectively managing their fertilizer usage. Moreover, there is a need for significant investment in breeding strategies aimed at developing new varieties of plants with enhanced nitrogen, phosphorus, and water use efficiency.

The complexity and global prevalence of agricultural systems can be addressed most effectively using transdisciplinary research. Here, we define transdisciplinary research that involves convergent and interdisciplinary methods to co-create knowledge related to societal problems among stakeholders and researchers (Scholz and Steiner 2015a, 2015b). Convergent research methods involve collaboration and knowledge exchange between interdisciplinary research groups. Transdisciplinary research is uniquely suited to addressing grand challenges, such as sustainable agricultural resource management, as it can foster collaboration, engage stakeholders, and generate practical solutions with real-world impact. In fact, transdisciplinary research has been used to address climate change challenges in certain regions around the world, including Europe, Africa, India, and Australia (de Jong et al. 2016; Serrao-Neumann et al. 2015; Vanderlinden et al. 2020). Expanding this approach in securing sustainable agricultural systems can help to meet the growing demands of the future.

If transdisciplinary research is to effectively address global challenges and coordinate meaningful solutions in society, more examples from biological and engineering applications are needed to demonstrate how it can be successfully implemented. Below, five case studies (Figure 2) are presented from two transdisciplinary and convergence research efforts. Ongoing efforts to advance sustainable and resilient agricultural systems involve integrative, multi-scale research that spans molecular, organismal, and ecosystem levels. This approach combines fundamental science with real-world applications to address global challenges in agricultural sustainability. Starting at the whole-plant scale, it is important to investigate how environmental inputs affect plant communication (signaling) and utilize synthetic biology approaches to harness this communication for reporting on resource availability. A focus on the immediate plant environment (soil) includes exploring new fertilizer technologies to increase the availability of phosphorus in the soil. At a plant population scale, developing tools for intimate interactions with plants is essential to identifying endophenotypes for improved breeding programs in the context of nitrogen and water use efficiency. Expanding to the ecosystem scale, investigating phosphorus recovery from aquatic systems and developing water resource management strategies are crucial to mitigating the harmful effects of phosphorus leaching. Finally, research is needed to understand stakeholder needs and priorities and to incorporate these findings back into research efforts. Showcasing these transdisciplinary endeavors highlights the importance of employing these approaches to tackle global challenges. By integrating diverse expertise and perspectives, more comprehensive solutions can be developed that address complex issues, fostering greater collaboration across scientific fields. Succinct explanations of technical terms used throughout the perspective can be found in the Supporting Information, Glossary.

2 Whole-Plant Reporting of Below-Ground Resources for Improved Crop Management

Plants simultaneously exist in, and draw resources from, multiple distinct environments. For example, plant roots are embedded in terrestrial landscapes, and their shoots inhabit the air. For whole-plant coordination and resource allocation between these environments, plants have the innate ability to send long-distance signals (Takahashi and Shinozaki 2019; Thomas and Frank 2019). These signals exist across scales: within cells (Woodson 2019), between cells (Lee 2015), from root-to-shoot (G. Wang et al. 2019), from leaf-to-leaf (Carmody et al. 2016), and between plants and their environment (Kegge and Pierik 2010). For example, CLE25 (CLAVATA3/EMBRYO-SURROUNDING REGION-RELATED 25) is a small peptide that moves through the xylem, from roots to shoots, to activate ABA-mediated drought stress responses in shoots (Takahashi et al. 2018). Another peptide involved in long-distance signals is CEP1 (C-TERMINALLY ENCODED PEPTIDE 1) that moves from rootstoshoots to activate nutrient foraging in response to low nitrogen (Ohkubo et al. 2017). MicroRNA2111 is a shoot-to-root small RNA signal that functions in balancing nodulation with rhizobial symbionts (Tsikou et al. 2018). Localized stress induces the rapid propagation of “waves” comprised of chemical (Bellandi et al. 2022; Toyota et al. 2018), electrical (Moe-Lange et al. 2021) and mechanical (Bacheva et al. 2024) responses that travel through plant tissue to induce systemic acquired acclimation (SAA) in distal tissues via the regulation of stomatal aperture and gene expression (Gilroy et al. 2016; Johns et al. 2021). By leveraging such diverse native long-distance signaling pathways and integrating approaches from plant science, synthetic biology, and evolutionary genomics (as shown in Figure 3), we can access information from otherwise inaccessible parts of the plant, such as roots, through reporters in accessible regions like the canopy. These reporter plants have the potential to revolutionize our understanding of plant communication and enable the efficient and timely application of resources (e.g., fertilizers, water). For example, a few reporter plants strategically placed in a field could be engineered to change leaf color or emit a fluorescent signal in response to water stress or nitrogen deficiency. Emission of such signals would enable farmers to target irrigation and fertilizer application precisely where needed, minimizing resource wastage.

In addition to modifying native signaling pathways, orthogonal pathways can be constructed (i.e., pathways that do not interfere with native pathways) to overcome undesired off-target effects of modifying native components. For example, synthetic reporters can be developed to indicate the activity of native or orthogonal pathways, and synthetic switches can be designed to toggle native or orthogonal pathways for a desired effect (e.g., stomatal closure, change in gene expression). Synthetic reporters have the potential to be multiplexed, enabling a single plant to report on multiple environmental stresses and improving resource management strategies. Synthetic switches can combine chemically induced genetic switches (Andres et al. 2019) with state-of-the-art optogenetic switches, the latter utilizing specific wavelengths of light to rapidly toggle gene expression on and off. Counterintuitively, most optogenetic tools are derived from plant signaling proteins (e.g., photoreceptors) but were initially used in bacteria (Lindner and Diepold 2022). Only recently have such optogenetic switches been introduced back into plants.

A challenge in modifying plant signaling pathways in crop species is that most knowledge of the networks underlying these pathways comes from the genetic model organism Arabidopsis thaliana, a small weed from the Brassicaceae (mustard) family. Studies in A. thaliana have significantly advanced the understanding of plant genomes, development, and function, but important challenges remain in translating these findings to crop species, some of which last shared a common ancestor with A. thaliana approximately 150 million years ago. To address this gap, each network or pathway is analyzed through an evolutionary framework (Figure 3—left). This workflow involves (i) inferring the phylogeny of each signaling pathway component, (ii) reconciling the recovered gene trees with established organismal relationships (termed gene tree-species tree reconciliation), and (ii) integrating functional and expression data with the reconciled trees to assess the extent of conservation, a form of evolutionary systems biology (O'Malley et al. 2015). This approach helps identify signaling pathways that are either highly conserved and suitable for engineering or variable across species, presenting opportunities to modify their effects when introduced outside their native species.

In addition to engineering plants for below-ground reporting, native microbiomes in the rhizosphere could also be harnessed for improved crop management. Growing evidence highlights the essential role of the rhizomicrobiome in facilitating plant nutrient uptake by modulating soil nutrient availability, stimulating plant growth, and activating defense mechanisms (Berendsen et al. 2012; Chaparro et al. 2014). Therefore, a better understanding of plant–microbe–soil interactions and the ability to harness these beneficial relationships could significantly reduce reliance on external inputs, such as synthetic fertilizers. Current “omics”-centric approaches, including genomics, transcriptomics, metabolomics, proteomics, and lipidomics, have provided extensive genotypic and phenotypic information on the rhizomicrobiome (White et al. 2017), (Bandara et al. 2021; Biswas and Sarkar 2018; Prosser 2015). However, these approaches lack the resolution to capture species-specific molecular interactions due to the largely unculturable and diverse nature of the rhizomicrobiome, spatiotemporal heterogeneity in soils and gene expression, and the loss of spatial and temporal cellular information (Bandara et al. 2021; Biswas and Sarkar 2018; Prosser 2015). Emerging single-cell technologies, such as Single Cell Raman microSpectroscopy (SCRS), offer a means to overcome some of these limitations by enabling in situ profiling of rhizomicrobiome communities at the single-cell level (Hatzenpichler et al. 2020; G. Li et al. 2022; H.-Z. Li et al. 2024; Y. Li et al. 2018; Majed et al. 2009; D. Wang et al. 2020, 2021; Z. Wang et al. 2023). Of particular interest are polyphosphate-accumulating organisms (PAOs), which could facilitate phosphorus uptake in plants, yet their roles in plant–microbe–soil interactions remain poorly understood (Akbari et al. 2021; X. Li et al. 2014; Mason-Jones et al. 2022; Zhao et al. 2021). Using SCRS, significant strides have been made in uncovering the dynamics of the abundant PAO community in the rhizosphere and identifying plant pathways linked to PAO activity (Baldwin et al. 2023, 2024; He et al. 2025).

3 Optimizing Novel Biomaterials for Improved Phosphorus Fertilization

An estimated 40% of the world's arable land is phosphorus limited (Zhu et al. 2018). This means that phosphorus or phosphate is either complexed with other molecules or is present in otherwise insoluble species, making most soil phosphorus largely bio-unavailable because inorganic orthophosphate is the only form available to plants (Madison et al. 2023; Zhu et al. 2018). Thus, phosphorus fertilizers are necessary for agricultural productivity because they supply inorganic orthophosphate, but they can contribute to environmental damage by leaching into aquatic ecosystems and promoting harmful algal blooms that are detrimental to wildlife (Shanmugavel et al. 2023). Accordingly, it will be necessary to find strategies to either make soil less phosphate-limited or to more sustainably fertilize crops. Biomaterials, such as functionalized hydrogels and nanofertilizers, are potential avenues to solve this problem, especially when co-created by chemists and biologists.

Functional hydrogels can be designed to cleave soluble, bioavailable phosphates from complexes, which would facilitate the goal of improving soil phosphate availability. Researchers have synthesized an organophosphate-degrading functionalized hydrogel by crosslinking poly (maleic anhydride-co-methyl vinyl ether) (PMAMVE) with ethylene diamine and reacting the remaining maleic anhydride units with hydroxylamine (Zboray et al. 2021). This process forms hydroxamic acid functional groups within the PMAMVE gel. The performance of the PMAMVE gels was evaluated in decomposing dimethyl nitrophenyl phosphate (DMNP), an organophosphate molecule. The decomposition of DMNP in the maleic anhydride gels followed pseudo-first-order kinetics for all studied conditions (Zboray et al. 2021). The spatial confinement of the hydroxamic acid groups inside the gel influenced the performance. The gels made of PMAMVE copolymers modified with hydroxamic acid offer a robust new system with high degradation efficiency, scalability, and simple preparation. Current efforts toward creating materials that degrade organophosphates involve modifying the PMAMVE gels with the deferoxamine (Desferal) and UiO-66 (Universitetet i Oslo) metal–organic framework. Once optimized by chemists and biologists, this hydrogel has the potential to improve phosphorus bioavailability from complex phosphate. Future steps toward optimization and implementation of this hydrogel will consider factors such as catalyst selection for optimal decomposition rates, organophosphate source, and soil pH. For example, the hydrogel will be made by crosslinking biocompatible poly (methyl vinyl ether-co-maleic anhydride) (PMVEMA) with polyethyleneimine (PEI) and then mixing them with an agar solution to form a substrate on which seeds can grow. The hydrogel is functionalized with imidazole and molybdate to catalyze the hydrolysis of organic phosphate, PNPP (p-nitrophenyl phosphate) instead of DMNP. The use of PNPP with the newly optimized hydrogel will be a superior system to DMNP and its decomposition with hydroxamic acid because it will operate at a pH range commonly found in soils (4.5–6.5) and decompose the phosphate at a slow rate. In this way, the rate of phosphate decomposition will occur at a slow rate and can be delivered in a substrate compatible with seeds.

A second avenue of designing novel fertilization methods involves nanofertilizers. Per IUPAC (International Union of Pure and Applied Chemistry), a nanoparticle is classified as a particle with a size less than 100 nm in any one of the three dimensions (Vert et al. 2012). Engineered nanoparticles often have new or novel physical or chemical properties compared to their bulk-scale counterparts. There are essentially two main methods to synthesize nanoparticles: top-down and bottom-up. Top-down synthesis involves grinding bulk material into nanoparticle size, while bottom-up synthesizes the material on the scale of nanometers using base or precursor materials. Top-down processes include different kinds of mechanical attrition using, for example, planetary or unitary ball mills. Bottom-up methods are built from molecular or atomic clusters to nucleate and grow the desired nanoparticle. Bottom-up processes include, for example, chemical precipitation, sol–gel method co-precipitation, hydrothermal synthesis, microemulsion, and electrochemical synthesis (Nagarajan 2008). To further enhance the properties of nanoparticles, they are often surface modified or functionalized, a process in which inorganic and organic molecules are added to their surface, depending on the covalent or polar nature of the nanoparticle (Ahmad et al. 2022; Thiruppathi et al. 2016).

In agriculture, functionalized nanoparticles are useful methods for obtaining controlled-release fertilizers to more sustainably utilize phosphate fertilizers. Nanoparticle controlled-release fertilizers exhibit better absorption and utilization of nutrients when compared to commercial fertilizers. Avila-Quezada et al. (2022) showed that nanofertilizers can reduce eutrophication due to the loss of nutrients, exhibit a higher diffusion and solubility rate, and have better controlled release properties when compared to commercially available controlled-release fertilizers.

Recent work in the field has found that macronutrients such as N, K, P, Ca, Mg, and S can be combined with nanomaterials to enable the delivery of a specific quantity of nutrients to crops (Nongbet et al. 2022). For example, Tarafdar et al. (2014) used phosphorus in tricalcium phosphate in concert with catalytic fungi to synthesize fungal-mediated phosphorus nanoparticles. This is an efficient way to maintain phosphorus in its usable form in the long term as well as allow the mobilization of phosphorus for plants (Nongbet et al. 2022). Moreover, Liu and Lal showed that treating soybean seeds with carboxymethyl cellulose-stabilized hydroxyapatite nanoparticles by ultrasonic dispersion resulted in the most effective fertilizer among those tested mainly because phosphorus demand is high at the local level and thus enhanced delivery at the nanoscale and in proximity facilitates proper absorption (Liu and Lal 2015). Interestingly, an important factor in nanoparticle controlled-release fertilizers is uniform shape. When the particles have consistently uniform size and shape, their contact angle with the plant will not change when fertilizer is applied, ultimately leading to better absorption (Sun et al. 2021).

In designing a nanofertilizer, a key parameter to consider is the fertilizer delivery method. If the application of the nanoparticles is foliar, then they are absorbed by trichomes, stomata, stigma, and hydathodes (Avila-Quezada et al. 2022). The nutrients are then transported through the phloem and xylem, where the nanoparticles enter the cells through endocytosis (Avila-Quezada et al. 2022; Lowry et al. 2024). It is currently understood that a target size for nanofertilizers of about a few tens of nm can lead to enhanced absorption by the plant (Sun et al. 2021). Nanofertilizers may also be designed to interact with or dissolve at the plant's outer surface, which requires consideration of a host of additional factors such as soil characteristics and plant cell wall chemistry (Lowry et al. 2024). Due to the physical, chemical, and biological constraints of nanofertilizer development, materials scientists, biologists, and social scientists can collaborate to optimize nanofertilizers to more efficiently and less wastefully deliver phosphate to plant cells. A wide array of factors must be considered, including economic, environmental, and biochemical factors. For example, nanoparticle environmental and ecological persistence and impacts over short and long-term periods must be evaluated along with hazard assessments, risk analyses, predictions of large-scale economic benefits and costs, and barriers to adoption in society (Avila-Quezada et al. 2022).

Nanofertilizer products have been made available on the market over the past decade (Yadav et al. 2023). Factors influencing bringing a nanofertilizer to market include an evaluation of its potential environmental impacts, efficacy of nutrient delivery to soil and plants, crop growth in response to the nanofertilizer, and an analysis of the costs of production and benefits of its commercialization (Yadav et al. 2023). Potential avenues for implementing nanofertilization methods include collaborations with university extension programs, particularly in land-grant universities, or pursuing industrial partnerships to perform initial evaluations of crop growth, soil nutrition, and toxicity in response to nanofertilizers. Interdisciplinary convergent research between biologists and economists and industry partnerships will also be key for economic analyses, at the production and utilization levels, to determine the viability of scale-up and commercialization of developed nanofertilizers.

Increased co-creation with the relevant stakeholders will address economic, regulatory, and social barriers to functionalized nanoparticles and hydrogels in agriculture. Further collaborations between biologists, chemists, and engineers will also optimize new fertilizer technologies like functionalized nanoparticles and hydrogels for practical application in field settings. Rapid screening pipelines will aid in iterative testing, refining, and co-creation of these technologies. 3D bioprinting has emerged as a technology that will enable rapid screening and testing of biomaterials such as hydrogels and nanoparticles. 3D bioprinting is an additive manufacturing technique in which protoplasts, or cells isolated by cell wall digestion, are extruded in a hydrogel with liquid media containing nutrients and hormones required to maintain cell viability for at least 2 weeks (Jose et al. 2016). In plants, 3D bioprinting has been optimized by researchers for Arabidopsis thaliana, soybean embryonic, shoot or root cells, and a variety of other plant cells (Madison et al. 2024, 2025; Van den Broeck et al. 2022). Cells are extruded by the bioprinter reproducibly and at a high throughput, making it a suitable platform for screening cellular responses to each biomaterial (Figure 4). Moreover, the 3D bioprinting techniques overcome current limitations of testing in protoplasts because protoplasts can be maintained for at least 2 weeks, allowing time for cell wall regrowth, cell division, and longer-term studies than is typical (Lowry et al. 2024). By using 3D bioprinting to test cell responses to each material, it will be easier and faster for researchers to optimize and refine each material for their compatibility with plant cells and improved cellular phosphorus content before testing and applying them to crops in field sites. It will also be possible to better refine each material by testing it with a complex combination of factors, such as pH levels, matrix types, and mineral compositions, that are closer approximations or models of field environments while still maintaining rapid, high throughput.

4 Scalable Field Endophenotyping for High-Throughput Screening of Plants

To accelerate advancements in breeding new cultivars for increased yield, nutritional value, resource efficiency, and resilience, it is necessary to develop technologies that enable high-throughput screening of plants and their associated organisms at molecular, tissue, and whole-organism scales across time and space in the field. This process of measuring plant processes generates large datasets that have been useful in building genomic predictions for crop improvements (Gage et al. 2019). Most plant phenotyping tasks have historically required manual effort, demanding a tedious and skilled workforce for completion (Atefi et al. 2021). Over the last decade, however, there has been a significant shift toward robotics and automation for high-throughput screening (Ruckelshausen et al. 2009). Efforts to estimate leaf area, color, shape, biomass, temperature, nutrient status, and plant growth have seen the introduction of technologies such as ground-based robots, aerial drones, and gantry systems. These systems are equipped with optical, hyperspectral, and thermal imaging as well as LIDAR (Light Detection and Ranging) and 3D scanning capabilities (Atefi et al. 2021; Qiao et al. 2022; Ruckelshausen et al. 2009; Wu et al. 2019). However, critical processes within the plant system remain inaccessible to interrogation and modulation for in situ and in-field and across diverse genetic backgrounds. For instance, current phenotyping efforts predominantly focus on visualization, leaving internal traits, or endophenotypes, largely unexplored. Moreover, while there is considerable focus on above-ground plant parts, below-ground aspects receive less attention due to technological constraints. Additionally, the existing robotic solutions are often rigid, whereas the nature of plant handling demands a delicate and gentle approach. To address these challenges, engineers and computer scientists are developing tools for in-field and high-throughput interactions with above- and below-ground components of plant systems, as shown in Figure 5.

Understanding water relations within plant tissue is necessary to enable the improvement of crop function relative to resilience and sustainability of water use. Leaf water potential has been shown to impact vegetative growth and yield, particularly during extreme drought conditions, making it a promising trait for improving water-use efficiency in plants. Specifically, the water potential of the living tissue between the xylem and the stomata plays a critical role in plant function but has been largely inaccessible due to a lack of tools. The recently developed nano-reporter (AquaDust) enables in situ and non-destructive measurement of water potential at this interface, providing unprecedented access to in-planta water status in a minimally invasive manner (Jain et al. 2021). The nano-reporter is infiltrated into the leaves and transduces the local water potential into a fluorescent spectral signal. AquaDust has been successfully used in maize (Jain et al. 2024a) and tomato (Jain et al. 2024b), showing its potential as an endophenotyping method for crop development and informing modeling efforts aimed at understanding water transport in plants. To enable automated and high-throughput endophenotyping using AquaDust, a soft-robotic system is being developed for automated leaf infiltration. This system consists of a soft gripper designed to autonomously infiltrate leaves with liquid solutions by gently squeezing them. In addition to enabling intimate interaction with the canopy, researchers are investigating approaches to improve below-ground phenotyping, which presents significant challenges due to the sensitive nature of the plant rhizosphere (Mishra, Tramacere, Guarino, et al. 2018; Mishra, Tramacere, and Mazzolai 2018). Accessing this zone without disrupting the soil ecosystem proves difficult, with very few existing solutions for underground manipulation (Bender et al. 2016). To address this, scalable underground soft robots are being developed—devices made from flexible, deformable materials such as silicone elastomers rather than rigid components—capable of accessing roots and the rhizosphere with minimal damage. Their compliant nature allows them to conform to and move through soil like earthworms, enabling direct connections with roots and the microbiome while preserving the delicate root architecture.

5 Phosphorus Recovery From Aquatic Ecosystems

Phosphorus can be present in many different forms (e.g., inorganic/organic, soluble/particulate) that dictate its behavior in water and soil matrices. In soils, many phosphorus species are present either as phosphate complexed with iron and other metal ions or are contained within organic molecules often referred to as legacy phosphorus (Briat et al. 2015; Madison et al. 2023; Poirier and Bucher 2002). Similarly, phosphorus is present as different species in aquatic systems. The soluble, reactive phosphorus fraction (orthophosphate) is directly available for biological uptake. Orthophosphate is also more readily recoverable using phosphorus treatment technologies applied in water and wastewater treatment systems, whereas the non-reactive phosphorus fraction is less studied. However, consideration of this fraction, which can comprise the majority of phosphorus in some water matrices, is important for sustainable phosphorus management. Collaborations between biologists and engineers can advance efforts to manage and utilize these ‘lost’ phosphorus fractions more efficiently. For example, different physical, chemical, and biological approaches can be used to facilitate phosphorus transformations to the target form.

In nature, phosphorus transformations occur at varying rates. Engineered systems can accelerate such processes to facilitate efficient phosphorus removal, recovery, and reuse in agriculture. For example, wastewater treatment facilities create conditions to remove phosphorus either chemically or biologically. Chemical transformation often involves adding iron or aluminum salts to precipitate phosphorus, through which soluble phosphate (e.g., $$$ {\mathrm{PO}}_4^{3-} $$$) is transformed into a solid form (e.g., AlPO_4(s)). Phosphorus transformation can also derive from physical–chemical processes, including oxidation processes. Redox reactions involve the transfer of electrons between chemical species. In water, phosphorus is typically present in the +5 valence state such that phosphorus bonds are not specifically targeted by oxidation. However, when phosphorus is incorporated in complex organic material, advanced oxidation processes may facilitate phosphorus recovery by partially oxidizing the phosphorus-containing materials into molecules more susceptible to hydrolytic conversion to phosphate or completely oxidizing the organic material to yield phosphate (Venkiteshwaran et al. 2018). A wide array of advanced oxidation processes have been tested as a means of breaking down complex or recalcitrant organics in water. These processes feature the generation of powerful oxidizing radical species such as hydroxyl radicals (OH·) capable of mineralizing organics to simple components such as $$$ {\mathrm{PO}}_4^{3-} $$$, water, and carbon dioxide.

Electrochemical processes are a type of advanced oxidation processes that use direct current to generate chemical reactions in situ by transferring electrons from an anode to a cathode (Figure 6a). This electrochemistry approach negates the need for transportation and storage of auxiliary chemicals. In electrooxidation, non-active electrode materials such as boron-doped diamond or Ti₄O₇ are often used to yield oxidation processes via direct oxidation on the electrode surface or indirect oxidation by hydroxyl radicals or other oxidants generated in situ (Ryan et al. 2021). Mallick et al. (2023) observed that electrooxidation featuring direct electron transfer was able to transform soluble non-reactive phosphorus to reactive forms, although the presence of organics in the water matrix interfered with process efficiency. The soluble reactive phosphorus compounds containing organic phosphoester P-O-C bonds were more susceptible to oxidation compared to inorganic anhydride P-O-P bonds (Mallick et al. 2021). Notably, complete conversion of non-reactive phosphorus to orthophosphate required high electrooxidation energy inputs such that incomplete conversion, when advanced oxidation is already in place as part of the water treatment regime, may be the most relevant means of implementation.

An example of biological transformation involves polyphosphate accumulating organisms (PAOs) in a process known as enhanced biological phosphorus removal (EBPR). PAOs take up soluble phosphate and transform it into intracellular chains of phosphate that are called polyphosphate (poly-P) (Figure 6b). PAOs are unique because they can hyperaccumulate phosphorus to levels exceeding 15% of a cell's total dry biomass. In water resource recovery facilities (WRRFs), the most common PAOs are Candidatus accumulibacter, Candidatus phosphoribacter (previously Tetrasphaera), and Azonexus (previously Dechloromonas) (Dueholm et al. 2022). EBPR depends on cycling between anaerobic and aerobic conditions to enrich PAOs. Under anaerobic conditions, PAOs consume carbon substrates such as volatile fatty acids, glucose, or amino acids and store them as polyhydroxyalkanoates (PHA), glycogen, or free amino acids (Kristiansen et al. 2013; McDaniel et al. 2021; Petriglieri et al. 2021; Singleton et al. 2022). Polyphosphate is a source of energy used to synthesize these carbon storage molecules, and its catabolism results in the release of inorganic orthophosphate from the cells. Under aerobic conditions, stored carbon is consumed to produce energy for cell growth and reproduction. In the process, polyphosphate stores are replenished, resulting in a net removal of phosphorus from wastewater. After settling, collecting, and treating the microbial biomass (called biosolids in the wastewater industry) that is rich in phosphorus, this biomass is often applied to agricultural lands. Biologically transformed phosphorus, such as poly-P, is more bioavailable to the soil and plants than chemically precipitated phosphorus. The intracellular poly-P is slowly degraded and released to the soil and plants (Figure 6b). If the settled biomass is treated in anaerobic digesters, the poly-P is released as orthophosphate, which can then be converted into field-ready fertilizers such as struvite. To our knowledge, outside of the wastewater treatment field, there are no other biotechnologies employing PAOs to remove phosphorus from aqueous matrices.

PAOs may play a role in phosphorus transformations in natural environments, including agricultural environments and those adjacent to agricultural settings (Akbari et al. 2021). Biologically derived poly-P is ubiquitous in nature though the contribution of specific PAOs in phosphorus cycling in natural environments is not well understood (Akbari et al. 2021; Saia et al. 2021). However, there is evidence of PAO-mediated phosphorus cycling in these environments. Saia et al. (2017) conducted ex situ experiments on native stream biofilms, demonstrating characteristic phosphorus uptake and release capabilities by these microbial communities. Coupled with evidence of intracellularly stored poly-P, Saia et al. demonstrated that poly-P accumulators in freshwater environments may contribute to phosphorus cycling, particularly during diel oxygenic conditions. Furthermore, Taylor et al. (2022) demonstrated that stream biofilms in nutrient-depleted sites stored more intracellular poly-P than those in nutrient-rich sites. Taylor et al. suggested that microbial communities in environments with higher legacy phosphorus store less poly-P. PAOs identified in EBPR have also been identified in sediments. Watson et al. (2019) detected Ca. accumulibacter in Columbia River estuary sediments that exhibited an increase in poly-P concentrations from dawn to dusk. Understanding the roles of PAOs for phosphorus transformations in natural environments and specifically how PAOs contribute to phosphorus cycling in agricultural soils is an opportunity for future research. By advancing understanding of both biological and physical–chemical phosphorus transformation processes, biologists and engineers can work together to better harness these powerful systems to improve phosphorus management by transforming phosphorus into forms that are more readily removed (to enhance protection of environmental waters from eutrophication), recovered (to encourage a circular phosphorus economy), and reused (to support global food production).

6 Integrating Stakeholder Perceptions and Needs Into Research and Innovation for Phosphorus Sustainability

Developing sustainable solutions for phosphorus management requires new technological innovations and improved management practices as well as a deeper understanding and integration of stakeholder needs and priorities (Brownlie et al. 2022; Grieger et al. 2024). By integrating stakeholder needs and priorities within research and innovation cycles, the proportion of technologies, management practices, and innovations (e.g., new crop varieties, smart fertilizers, and removal/recovery systems) that will ultimately be adopted by end-users increases substantially. In addition, better understanding and incorporating stakeholder needs and priorities within early phases of research and innovation align with broader concepts of responsible innovation that seek to better align scientific research and innovation efforts with societal needs and wants (Grieger et al. 2022; Stilgoe et al. 2013).

For these aforementioned reasons, conducting research to identify the perceptions and needs of diverse stakeholders involved in phosphorus management and sustainability using a range of approaches (Grieger, Barry, et al. 2025). Among other examples, a research team conducted a survey to understand the perceptions, needs, and potential concerns of expert stakeholders for achieving phosphorus sustainability (Grieger et al. 2022). Building on the results of the survey, targeted interviews were conducted with key stakeholders involved in projects that aimed to develop new technologies or strategies to capture phosphorus (A. W. Merck, Deaver, et al. 2024).

In more detail, the research team developed a survey in the fall of 2022 to investigate stakeholder views and needs for achieving phosphorus sustainability using the online platform Qualtrics. The survey was composed of 14 multiple-choice and open-ended questions and was deployed to expert stakeholders in the U.S. and abroad using email and posted on social media (LinkedIn) through the sustainable phosphorus alliance. Based on responses from 96 survey participants who provided complete and valid responses, results indicated that stakeholders largely considered phosphorus management to be unsustainable and they were concerned about the current ability to manage phosphorus in a sustainable way (Grieger et al. 2022). Reasons behind these views were based on the fact that phosphorus is mined from nonrenewable phosphate rock, there are significant challenges to phosphorus recovery and reuse, phosphorus management in agriculture is inefficient, and the environmental impacts are associated with the inefficient management of phosphorus. This study also found that stakeholders reported a range of needs to achieve phosphorus sustainability, including better management practices, new technologies, new or better regulations, and ways in which to engage other stakeholders. While these findings are not entirely surprising for researchers and practitioners working in phosphorus sustainability fields, this study was the first to document stakeholder needs and priorities to achieve phosphorus sustainability using a structured and rigorous approach based in social sciences (Grieger et al. 2022).

Next, the research team identified three case studies that represented promising technologies or new strategies to capture and recover phosphorus to improve sustainability. Building on the results of the previously described stakeholder survey, they investigated stakeholder perceptions and needs related to three case studies to better understand stakeholder views, needs, as well as perceived driving forces and barriers to technology acceptance and adoption related to these cases. The case studies selected were (i) phosphorus recovery from urine in buildings, (ii) improving EBPR in wastewater treatment, and (iii) isotopic tracking in natural waters to better understand the attribution of phosphorus pollution. A total of 37 interviews were conducted with key stakeholders involved in the case studies in the spring of 2023 to better understand their perceptions, needs, and potential concerns for achieving phosphorus sustainability (A. W. Merck, Deaver, et al. 2024). Key findings from this work include a finding that stakeholders identified a common set of economic, regulatory, and societal issues as main factors that affect the adoption of new scientific or technical solutions for phosphorus recovery. For example, the cost of water and phosphate rock were identified as primary drivers in the adoption of urine diversion in buildings, whereas technical improvements were a primary driver in the EBPR case. At the same time, barriers to technology adoption included technical, regulatory, and economic challenges, such as societal push-back associated with negative perceptions of urine diverting toilets in the urine case or technical challenges to EBPR.

In fact, policy and regulatory-focused challenges were considered to be a major barrier to the urine diversion case study based on interviews with key stakeholders and subsequent research that investigated perceptions of urine diversion technologies in academic buildings in the US according to university leaders and administrators (Grieger, Scholz, et al. 2025). Based on the results of these stakeholder studies, the research team co-created and published a policy paper that details the need to address regulatory barriers to scaling up urine diversion systems in the U.S. (A. Merck, Grieger, et al. 2024). Findings from this work are being integrated back into other research projects to improve both technical aspects and engagement efforts associated with these cases (Crane et al. 2024). Following these efforts, the research team is also developing collaborative working groups within specific technical projects to co-create solutions with stakeholders in specific contexts and settings as well as developing best practices and guidance to follow for other researchers who aim to engage stakeholders in their work related to phosphorus sustainability (Grieger, Scholz, et al. 2025).

In addition to understanding stakeholder perceptions, needs, and priorities, it is also important to consider research and innovation within larger regulatory and policy contexts, as mentioned above. Regulatory and policy contexts will likely be case-dependent, as research and innovation in food and agricultural contexts, for instance, will operate within separate sets of regulations and policies compared to, for example, EBPR relevant for WRRFs or implementing urine diversion systems in U.S. buildings. The importance of having effective regulations and policies to achieve phosphorus sustainability was also emphasized during the expert stakeholder survey conducted in 2022, in which new or improved regulations or policies were cited as the 3rd most important needs of stakeholders to achieve sustainability (Grieger et al. 2024). Therefore, transdisciplinary collaboration to improve sustainability will need to consider the broader regulatory and political landscapes in which research, innovation, and partnerships occur. By drawing on diverse fields such as natural science, engineering, social science, and economics, transdisciplinary research can help inform evidence-based policies, for example, through policy briefs and outreach. Likewise, policies and regulations can also influence research priorities and agenda-setting for transdisciplinary research collaborations.

Overall, these examples highlight how transdisciplinary and convergence research can be conducted in projects focused on specific aspects of phosphorus sustainability, through the integration of natural sciences, environmental engineering, and social sciences.

7 Integrating Stakeholder Perceptions for Nitrogen and Water Management

To integrate stakeholder perspectives into nitrogen and water management technologies, researchers should conduct both external engagement (surveys and interviews) and internal evaluation (within the research team) to ensure that innovations align with the practical realities of rural environments where technologies will be deployed. In particular, looking both early and inward in transdisciplinary research involving computing is important because prior literature indicates that computer systems developers inadvertently embed their own values into computing technologies—a term widely known as values in design (Nathan et al. 2008; Nissenbaum 2001; Shilton 2010).

Following this approach, in a recent study focusing on integrating stakeholder perspectives in computing technologies for nitrogen and water management, the team conducted a study with two components. (i) A 24-month autobiographical study where researchers deployed experimental networking technologies for scalable field computing while critically reflecting on their day-to-day systems building to anticipate potential societal impacts of early technical design decisions. (ii) A set of seven semi-structured interviews with researchers and farmers involved in early phase adoption and adaptation of similar networked technologies for data-driven agriculture to understand their challenges, attitudes, and visions of the emerging technologies. Key findings from the study include a 100 year gap between the seamless visions of researchers and the seamful realities under which the technologies are being built and expected to operate (Rubambiza et al. 2022). For example, one participant described realities where “some of my neighbors farm like it is 1950”. Furthermore, as a result of the environments under which research is conducted (e.g., ample university IT support and research funding), researchers maintain faith in an eventual seamlessness of the technologies in the contexts where they will be deployed (i.e., rural farms).

The century gap is crucial to acknowledge because it could affect consumer adoption of technologies built in the lab to address global challenges but not grounded in the day-to-day realities of farms. More importantly, the findings raised three implications for the ongoing design, implementation, and evaluation of scalable field computing technologies within this research: (i) supporting the right to tinker with the technologies to fit the particular environment where they could be deployed; (ii) documenting invisible work (e.g., hardware and cloud computing service choices) and their effects on the design decisions; and (iii) bridging the gulf to farms by recognizing farmers as natural tinkerers who have much insight to offer to technology designs and implementation.

8 Expanding Towards Global Applications of Transdisciplinary Research

The challenges in agriculture, along with the scientific and technological approaches introduced above, all have global relevance. Pressures on resources such as phosphorus span international borders, as do the downstream impacts of inefficient use of resources on natural ecosystems. The development and translation of new technologies such as nanofertilizers and reporter plants could help mitigate these challenges internationally. To achieve this goal requires intentional engagement with colleagues around the globe, with a recognition that potentially distinct cultural dimensions of collaboration can add to those of a transdisciplinary effort (Werlen 2015). Economic circumstances and policy contexts internationally also present important additional considerations, particularly with respect to evolving perceptions and regulations on emerging nanotechnologies (Kumari et al. 2023) and bioengineered plants (Entine et al. 2021).

9 Conclusions

With a changing climate, a growing population, and decreasing availability of resources such as water, phosphorus, and nitrogen, there is an urgent need to improve the productivity and sustainability of crop-based agriculture. Transdisciplinary research has emerged as a research model that can address the coupled scientific, technological, and societal dimensions of this challenge in a manner that defines successful convergence research. In this paper, we highlight several case studies concerning the implementation of transdisciplinary and convergence research, including research to develop reporter plants of below-ground nutrient and water availability as well as the integration of plant-microbial interactions as efficient strategies for agricultural resource management. Moreover, plant-informed early optimization of new phosphorus fertilization technologies is in development to help reduce the overuse of applied phosphorus fertilizer by either improving soil phosphorus bioavailability or implementing more efficient nanofertilizer methods. Integrating nanotechnology with robotics, we show how we can accelerate advancements in breeding new cultivars. In addition, phosphorus recovery strategies are being developed using electrooxidation or polyphosphate-accumulating organism systems to help resolve the problem of phosphorus contamination of aquatic ecosystems. Finally, it is important to identify stakeholder needs and barriers to the adoption of research projects to guide institutions in advancing work that effectively supports their target societal goals.

As transdisciplinary research is implemented in future applications, it will be important to consider a number of factors such as diversity in collaborations, cross-training among disciplines, and co-creation with stakeholders (Scholz and Steiner 2015b; Werlen 2015). Transdisciplinary convergence research efficiently facilitates innovation as each team can combine their expertise in new ways that, in isolation, they would not be able to implement in a timely manner, if at all. For example, endophenotyping enabled by nanoreporters and robotics relies on collaboration between engineers and plant biologists, while fertilizer biomaterial optimization relies on collaboration between chemists, materials scientists, and plant biologists. Both of these developments provide new opportunities to engage stakeholders in the design of application-appropriate research questions and technologies. Seeking out opportunities to co-create in multidisciplinary teams will thus help bring innovative ideas to fruition without the need for the impossible task of one individual team building expertise in all disciplines. Furthermore, innovation and technology do not exist in a vacuum. Social factors such as stakeholder needs and perceptions, governmental regulations, economic factors, environmental risks, and practicality of using the technology all need to be considered (Cundill et al. 2019; Scholz and Steiner 2015a, 2015b). To do this, stakeholder engagement and co-creation are key to guiding the focus of transdisciplinary research projects in real time. Accordingly, collaboration with researchers within the social sciences is necessary to produce meaningful innovations. Thus, transdisciplinary research implemented to its full potential brings efficiency and direction to research and innovation towards solving complex global challenges.

Author Contributions

Vesna Bacheva: conceptualization, writing – original draft, writing – review and editing. Imani Madison: conceptualization, writing – original draft, writing – review and editing. Mathew Baldwin: writing – review and editing. Justin Baker: writing – review and editing. Mark Beilstein: writing – review and editing. Douglas F. Call: writing – review and editing. Jessica A. Deaver: writing – review and editing. Kirill Efimenko: writing – review and editing. Jan Genzer: writing – review and editing. Khara Grieger: writing – review and editing. April Z. Gu: writing – review and editing. Mehmet Mert Ilman: writing – review and editing. Jen Liu: writing – review and editing. Sijin Li: writing – original draft. Schmidt Science Fellows, in partnership with the Rhodes Trust. Brooke K. Mayer: writing – review and editing. Anand Kumar Mishra: writing – review and editing. Juan Claudio Nino: writing – review and editing. Gloire Rubambiza: writing – review and editing. Phoebe Sengers: writing – review and editing. Robert Shepherd: writing – review and editing. Jesse Woodson: writing – review and editing. Hakim Weatherspoon: writing – review and editing. Margaret Frank: conceptualization, writing – original draft, writing – review and editing. Jacob L. Jones: conceptualization, writing – original draft, writing – review and editing. Rosangela Sozzani: conceptualization, writing – original draft, writing – review and editing. Abraham D. Stroock: conceptualization, writing – original draft, writing – review and editing.

Conflicts of Interest

Rosangela Sozzani declares a conflicts of interest with Raleigh BioSciences. The other authors declare no conflicts of interest.