Spectral differentiation of oak wilt from foliar fungal disease and drought is correlated with physiological changes

Machine learning in the classification of asian rust severity in soybean using hyperspectral sensor

Traditional monitoring of asian soybean rust severity is a time- and labor-intensive task, as it requires visual assessments by skilled professionals in the field. Thus, the use of remote sensing and machine learning (ML) techniques in data processing has emerged as an approach that can increase efficiency in disease monitoring, enabling faster, more accurate and time- and labor-saving evaluations. The aims of the study were: (i) to identify the spectral signature of different levels of Asian soybean rust severity; (ii) to identify the most accurate machine learning algorithm for classifying disease severity levels; (iii) which spectral input provides the highest classification accuracy for the algorithms; (iv) to determine a sample size of leaves that guarantees the best accuracy for the algorithms. A field experiment was carried out in the 2022/2023 harvest in a randomized block design with a 6x3 factorial scheme (ML algorithms x severity levels) and four replications. Disease severity levels assessed were: healthy leaves, 25 % severity, and 50 % severity. Leaf hyperspectral analysis was carried out over a wide range from 350 to 2500 nm. From this analysis, 28 spectral bands were extracted, seeking to distinguish the spectral signature for each severity level with the least input dataset. Data was subjected to machine learning analysis using Artificial Neural Network (ANN), REPTree (DT) and J48 decision trees, Random Forest (RF), and Support Vector Machine (SVM) algorithms, as well as a traditional classification method (Logistic Regression - LR). Two different input datasets were tested for each algorithm: the full spectrum (ALL) provided by the sensor and the 28 spectral bands (SB). Tests with different sample sizes were also conducted to investigate the algorithms' ability to detect severity levels with a reduced sample size. Our findings indicate differences between the spectral curves for the severity levels assessed, which makes it possible to differentiate between healthy plants with low and high severity using hyperspectral sensing. SVM was the most accurate algorithm for classifying severity levels by using all the spectral information as input. This algorithm also provided high classification accuracy when using smaller leaf samples. This study reveals that hyperspectral sensing and the use of ML algorithms provide an accurate classification of different levels of Asian rust severity, and can be powerful tools for a more efficient disease monitoring process.

Foliar spectra accurately distinguish most temperate tree species and show strong phylogenetic signal

PREMISE

Spectroscopy is a powerful remote sensing tool for monitoring plant biodiversity over broad geographic areas. Increasing evidence suggests that foliar spectral reflectance can be used to identify trees at the species level. However, most studies have focused on only a limited number of species at a time, and few studies have explored the underlying phylogenetic structure of leaf spectra. Accurate species identifications are important for reliable estimations of biodiversity from spectral data.

METHODS

Using over 3500 leaf-level spectral measurements, we evaluated whether foliar reflectance spectra (400-2400 nm) can accurately differentiate most tree species from a regional species pool in eastern North America. We explored relationships between spectral, phylogenetic, and leaf functional trait variation as well as their influence on species classification using a hurdle regression model.

RESULTS

Spectral reflectance accurately differentiated tree species (κ = 0.736, ±0.005). Foliar spectra showed strong phylogenetic signal, and classification errors from foliar spectra, although present at higher taxonomic levels, were found predominantly between closely related species, often of the same genus. In addition, we find functional and phylogenetic distance broadly control the occurrence and frequency of spectral classification mistakes among species.

CONCLUSIONS

Our results further support the link between leaf spectral diversity, taxonomic hierarchy, and phylogenetic and functional diversity, and highlight the potential of spectroscopy to remotely sense plant biodiversity and vegetation response to global change.

Mechanistic links between physiology and spectral reflectance enable previsual detection of oak wilt and drought stress

Tree mortality due to global change-including range expansion of invasive pests and pathogens-is a paramount threat to forest ecosystems. Oak forests are among the most prevalent and valuable ecosystems both ecologically and economically in the United States. There is increasing interest in monitoring oak decline and death due to both drought and the oak wilt pathogen (Bretziella fagacearum). We combined anatomical and ecophysiological measurements with spectroscopy at leaf, canopy, and airborne levels to enable differentiation of oak wilt and drought, and detection prior to visible symptom appearance. We performed an outdoor potted experiment with Quercus rubra saplings subjected to drought stress and/or artificially inoculated with the pathogen. Models developed from spectral reflectance accurately predicted ecophysiological indicators of oak wilt and drought decline in both potted and field experiments with naturally grown saplings. Both oak wilt and drought resulted in blocked water transport through xylem conduits. However, oak wilt impaired conduits in localized regions of the xylem due to formation of tyloses instead of emboli. The localized tylose formation resulted in more variable canopy photosynthesis and water content in diseased trees than drought-stressed ones. Reflectance signatures of plant photosynthesis, water content, and cellular damage detected oak wilt and drought 12 d before visual symptoms appeared. Our results show that leaf spectral reflectance models predict ecophysiological processes relevant to detection and differentiation of disease and drought. Coupling spectral models that detect physiological change with spatial information enhances capacity to differentiate plant stress types such as oak wilt and drought.

Proximal hyperspectral sensing of abiotic stresses in plants

Recent attempts, advances and challenges, as well as future perspectives regarding the application of proximal hyperspectral sensing (where sensors are placed within 10 m above plants, either on land-based platforms or in controlled environments) to assess plant abiotic stresses have been critically reviewed. Abiotic stresses, caused by either physical or chemical reasons such as nutrient deficiency, drought, salinity, heavy metals, herbicides, extreme temperatures, and so on, may be more damaging than biotic stresses (affected by infectious agents such as bacteria, fungi, insects, etc.) on crop yields. The proximal hyperspectral sensing provides images at a sub-millimeter spatial resolution for doing an in-depth study of plant physiology and thus offers a global view of the plant's status and allows for monitoring spatio-temporal variations from large geographical areas reliably and economically. The literature update has been based on 362 research papers in this field, published from 2010, most of which are from four years ago and, in our knowledge, it is the first paper that provides a comprehensive review of the applications of the technique for the detection of various types of abiotic stresses in plants.

Exploring plant responses to abiotic stress by contrasting spectral signature changes

In this study, daily changes over a short period and diurnal progression of spectral reflectance at the leaf level were used to identify spring wheat genotypes (Triticum aestivum L.) susceptible to adverse conditions. Four genotypes were grown in pots experiments under semi-controlled conditions in Chile and Spain. Three treatments were applied: i) control (C), ii) water stress (WS), and iii) combined water and heat shock (WS+T). Spectral reflectance, gas exchange and chlorophyll fluorescence measurements were performed on flag leaves for three consecutive days at anthesis. High canopy temperature ( H _CT ) genotypes showed less variability in their mean spectral reflectance signature and chlorophyll fluorescence, which was related to weaker responses to environmental fluctuations. While low canopy temperature ( L _CT ) genotypes showed greater variability. The genotypes spectral signature changes, in accordance with environmental fluctuation, were associated with variations in their stomatal conductance under both stress conditions (WS and WS+T); L _CT genotypes showed an anisohydric response compared that of H _CT , which was isohydric. This approach could be used in breeding programs for screening a large number of genotypes through proximal or remote sensing tools and be a novel but simple way to identify groups of genotypes with contrasting performances.

Multi-sensor spectral synergies for crop stress detection and monitoring in the optical domain: A review

Remote detection and monitoring of the vegetation responses to stress became relevant for sustainable agriculture. Ongoing developments in optical remote sensing technologies have provided tools to increase our understanding of stress-related physiological processes. Therefore, this study aimed to provide an overview of the main spectral technologies and retrieval approaches for detecting crop stress in agriculture. Firstly, we present integrated views on: i) biotic and abiotic stress factors, the phases of stress, and respective plant responses, and ii) the affected traits, appropriate spectral domains and corresponding methods for measuring traits remotely. Secondly, representative results of a systematic literature analysis are highlighted, identifying the current status and possible future trends in stress detection and monitoring. Distinct plant responses occurring under shortterm, medium-term or severe chronic stress exposure can be captured with remote sensing due to specific light interaction processes, such as absorption and scattering manifested in the reflected radiance, i.e. visible (VIS), near infrared (NIR), shortwave infrared, and emitted radiance, i.e. solar-induced fluorescence and thermal infrared (TIR). From the analysis of 96 research papers, the following trends can be observed: increasing usage of satellite and unmanned aerial vehicle data in parallel with a shift in methods from simpler parametric approaches towards more advanced physically-based and hybrid models. Most study designs were largely driven by sensor availability and practical economic reasons, leading to the common usage of VIS-NIR-TIR sensor combinations. The majority of reviewed studies compared stress proxies calculated from single-source sensor domains rather than using data in a synergistic way. We identified new ways forward as guidance for improved synergistic usage of spectral domains for stress detection: (1) combined acquisition of data from multiple sensors for analysing multiple stress responses simultaneously (holistic view); (2) simultaneous retrieval of plant traits combining multi-domain radiative transfer models and machine learning methods; (3) assimilation of estimated plant traits from distinct spectral domains into integrated crop growth models. As a future outlook, we recommend combining multiple remote sensing data streams into crop model assimilation schemes to build up Digital Twins of agroecosystems, which may provide the most efficient way to detect the diversity of environmental and biotic stresses and thus enable respective management decisions.

Digital plant pathology: a foundation and guide to modern agriculture

Over the last 20 years, researchers in the field of digital plant pathology have chased the goal to implement sensors, machine learning and new technologies into knowledge-based methods for plant phenotyping and plant protection. However, the application of swiftly developing technologies has posed many challenges. Greenhouse and field applications are complex and differ in their study design requirements. Selecting a sensor type (e.g., thermography or hyperspectral imaging), sensor platform (e.g., rovers, unmanned aerial vehicles, or satellites), and the problem-specific spatial and temporal scale adds to the challenge as all pathosystems are unique and differ in their interactions and symptoms, or lack thereof. Adding host-pathogen-environment interactions across time and space increases the complexity even further. Large data sets are necessary to enable a deeper understanding of these interactions. Therefore, modern machine learning methods are developed to realize the fast data analysis of such complex data sets. This reduces not only human effort but also enables an objective data perusal. Especially deep learning approaches show a high potential to identify probable cohesive parameters during plant-pathogen-environment interactions. Unfortunately, the performance and reliability of developed methods are often doubted by the potential user. Gaining their trust is thus needed for real field applications. Linking biological causes to machine learning features and a clear communication, even for non-experts of such results, is a crucial task that will bridge the gap between theory and praxis of a newly developed application. Therefore, we suggest a global connection of experts and data as the basis for defining a common and goal-oriented research roadmap. Such high interconnectivity will likely increase the chances of swift, successful progress in research and practice. A coordination within international excellence clusters will be useful to reduce redundancy of research while supporting the creation and progress of complementary research. With this review, we would like to discuss past research, achievements, as well as recurring and new challenges. Having such a retrospect available, we will attempt to reveal future challenges and provide a possible direction elevating the next decade of research in digital plant pathology.

Early detection of a tree pathogen using airborne remote sensing

Native forests of Hawai'i Island are experiencing an ecological crisis in the form of Rapid 'Ōhi'a Death (ROD), a recently characterized disease caused by two fungal pathogens in the genus Ceratocystis. Since approximately 2010, this disease has caused extensive mortality of Hawai'i's keystone endemic tree, known as 'ōhi'a (Metrosideros polymorpha). Visible symptoms of ROD include rapid browning of canopy leaves, followed by death of the tree within weeks. This quick progression leading to tree mortality makes early detection critical to understanding where the disease will move at a timescale feasible for controlling the disease. We used repeat laser-guided imaging spectroscopy (LGIS) of forests on Hawai'i Island collected by the Global Airborne Observatory (GAO) in 2018 and 2019 to derive maps of foliar trait indices previously found to be important in distinguishing between ROD-infected and healthy 'ōhi'a canopies. Data from these maps were used to develop a prognostic indicator of tree stress prior to the visible onset of browning. We identified canopies that were green in 2018, but became brown in 2019 (defined as "to become brown"; TBB), and a corresponding set of canopies that remained green. The data mapped in 2018 showed separability of foliar trait indices between TBB and green 'ōhi'a, indicating early detection of canopy stress prior to the onset of ROD. Overall, a combination of linear and non-linear analyses revealed canopy water content (CWC), foliar tannins, leaf mass per area (LMA), phenols, cellulose, and non-structural carbohydrates (NSC) are primary drivers of the prognostic spectral capability which collectively result in strong consistent changes in leaf spectral reflectance in the near-infrared (700-1300 nm) and shortwave-infrared regions (1300-2500 nm). Results provide insight into the underlying foliar traits that are indicative of physiological responses of M. polymorpha trees infected with Ceratocycstis and suggest that imaging spectroscopy is an effective tool for identifying trees likely to succumb to ROD prior to the onset of visible symptoms.

Plant Disease Sensing: Studying Plant-Pathogen Interactions at Scale

Plant disease threatens the environmental and financial sustainability of crop production, causing $220 billion in annual losses. The dire threat disease poses to modern agriculture demands tools for better detection and monitoring to prevent crop loss and input waste. The nascent discipline of plant disease sensing, or the science of using proximal and/or remote sensing to detect and diagnose disease, offers great promise to extend monitoring to previously unachievable resolutions, a basis to construct multiscale surveillance networks for early warning, alert, and response at low latency, an opportunity to mitigate loss while optimizing protection, and a dynamic new dimension to agricultural systems biology. Despite its revolutionary potential, plant disease sensing remains an underdeveloped discipline, with challenges facing both fundamental study and field application. This article offers a perspective on the current state and future of plant disease sensing, highlights remaining gaps to be filled, and presents a bold vision for the future of global agriculture.

Grapevines under drought do not express esca leaf symptoms

In the context of climate change, plant mortality is increasing worldwide in both natural and agroecosystems. However, our understanding of the underlying causes is limited by the complex interactions between abiotic and biotic factors and the technical challenges that limit investigations of these interactions. Here, we studied the interaction between two main drivers of mortality, drought and vascular disease (esca), in one of the world's most economically valuable fruit crops, grapevine. We found that drought totally inhibited esca leaf symptom expression. We disentangled the plant physiological response to the two stresses by quantifying whole-plant water relations (i.e., water potential and stomatal conductance) and carbon balance (i.e., CO₂ assimilation, chlorophyll, and nonstructural carbohydrates). Our results highlight the distinct physiology behind these two stress responses, indicating that esca (and subsequent stomatal conductance decline) does not result from decreases in water potential and generates different gas exchange and nonstructural carbohydrate seasonal dynamics compared to drought.

Detection of the metabolic response to drought stress using hyperspectral reflectance

Drought is the most important limitation on crop yield. Understanding and detecting drought stress in crops is vital for improving water use efficiency through effective breeding and management. Leaf reflectance spectroscopy offers a rapid, non-destructive alternative to traditional techniques for measuring plant traits involved in a drought response. We measured drought stress in six glasshouse-grown agronomic species using physiological, biochemical, and spectral data. In contrast to physiological traits, leaf metabolite concentrations revealed drought stress before it was visible to the naked eye. We used full-spectrum leaf reflectance data to predict metabolite concentrations using partial least-squares regression, with validation R2 values of 0.49-0.87. We show for the first time that spectroscopy may be used for the quantitative estimation of proline and abscisic acid, demonstrating the first use of hyperspectral data to detect a phytohormone. We used linear discriminant analysis and partial least squares discriminant analysis to differentiate between watered plants and those subjected to drought based on measured traits (accuracy: 71%) and raw spectral data (66%). Finally, we validated our glasshouse-developed models in an independent field trial. We demonstrate that spectroscopy can detect drought stress via underlying biochemical changes, before visual differences occur, representing a powerful advance for measuring limitations on yield.

Monitoring natural and anthropogenic plant stressors by hyperspectral remote sensing: Recommendations and guidelines based on a meta-review

This review outlines the advances achieved in monitoring natural and anthropogenic plant stressors by hyperspectral remote sensing over the last 50 years. A broad diversity of methods based on field and imaging spectroscopy were developed in that field for precision farming and environmental monitoring purposes. From the 466 articles reviewed, we identified the main factors to consider to achieve accurate monitoring of plant stress, namely: The plant species and the stressor to monitor, the goal (detection or quantification), and scale (field or broad-scale) of monitoring, and the need for controlled experiments. Based on these factors, we then provide recommendations and guidelines for the development of reliable methods to monitor 11 major biotic and abiotic plant stressors. For each stressor, the effects on plant health and reflectance are described and the most suited spectral regions, scale, spatial resolution, and processing approaches to achieve accurate monitoring are presented. As a perspective, we discuss two major components that should be implemented in future methods to improve stress monitoring: The discrimination of plant stressors with similar effects on plants and the transferability of the methods across scales.

Spatial Patterns of ‘Ōhi‘a Mortality Associated with Rapid ‘Ōhi‘a Death and Ungulate Presence

Effective forest management, particularly during forest disturbance events, requires timely and accurate monitoring information at appropriate spatial scales. In Hawai‘i, widespread ‘ōhi‘a (Metrosideros polymorpha Gaud.) mortality associated with introduced fungal pathogens affects forest stands across the archipelago, further impacting native ecosystems already under threat from invasive species. Here, we share results from an integrated monitoring program based on high resolution (<5 cm) aerial imagery, field sampling, and confirmatory laboratory testing to detect and monitor ‘ōhi‘a mortality at the individual tree level across four representative sites on Hawai‘i island. We developed a custom imaging system for helicopter operations to map thousands of hectares (ha) per flight, a more useful scale than the ten to hundreds of ha typically covered using small, unoccupied aerial systems. Based on collected imagery, we developed a rating system of canopy condition to identify ‘ōhi‘a trees suspected of infection by the fungal pathogens responsible for rapid ‘ōhi‘a death (ROD); we used this system to quickly generate and share suspect tree candidate locations with partner agencies to rapidly detect new mortality outbreaks and prioritize field sampling efforts. In three of the four sites, 98% of laboratory samples collected from suspect trees assigned a high confidence rating (n = 50) and 89% of those assigned a medium confidence rating (n = 117) returned positive detections for the fungal pathogens responsible for ROD. The fourth site, which has a history of unexplained ‘ōhi‘a mortality, exhibited much lower positive detection rates: only 6% of sampled trees assigned a high confidence rating (n = 16) and 0% of the sampled suspect trees assigned a medium confidence rating (n = 20) were found to be positive for the pathogen. The disparity in positive detection rates across study sites illustrates challenges to definitively determine the cause of ‘ōhi‘a mortality from aerial imagery alone. Spatial patterns of ROD-associated ‘ōhi‘a mortality were strongly affected by ungulate presence or absence as measured by the density of suspected ROD trees in fenced (i.e., ungulate-free) and unfenced (i.e., ungulate present) areas. Suspected ROD tree densities in neighboring areas containing ungulates were two to 69 times greater than those found in ungulate-free zones. In one study site, a fence line breach occurred during the study period, and feral ungulates entered an area that was previously ungulate-free. Following the breach, suspect ROD tree densities in this area rose from 0.02 to 2.78 suspect trees/ha, highlighting the need for ungulate control to protect ‘ōhi‘a stands from Ceratocystis-induced mortality and repeat monitoring to detect forest changes and resource threats.

Plant pest surveillance: from satellites to molecules

Plant pests and diseases impact both food security and natural ecosystems, and the impact has been accelerated in recent years due to several confounding factors. The globalisation of trade has moved pests out of natural ranges, creating damaging epidemics in new regions. Climate change has extended the range of pests and the pathogens they vector. Resistance to agrochemicals has made pathogens, pests, and weeds more difficult to control. Early detection is critical to achieve effective control, both from a biosecurity as well as an endemic pest perspective. Molecular diagnostics has revolutionised our ability to identify pests and diseases over the past two decades, but more recent technological innovations are enabling us to achieve better pest surveillance. In this review, we will explore the different technologies that are enabling this advancing capability and discuss the drivers that will shape its future deployment.

Towards In Vivo Monitoring of Ions Accumulation in Trees: Response of an in Planta Organic Electrochemical Transistor Based Sensor to Water Flux Density, Light and Vapor Pressure Deficit Variation

Research on organic electrochemical transistor (OECT) based sensors to monitor in vivo plant traits such as xylem sap concentration is attracting attention for their potential application in precision agriculture. Fabrication and electronic aspects of OECT have been the subject of extensive research while its characterization within the plant water relation context deserves further efforts. This study tested the hypothesis that the response (R) of an OECT (bioristor) implanted in the trunk of olive trees is inversely proportional to the water flux density flowing through the plant (Jw). This study also examined the influence on R of vapor pressure deficit (VPD) as coupled/uncoupled with light. R was hourly recorded in potted olive trees for a 10-day period concomitantly with Jw (weight loss method). A subgroup of trees was bagged in order to reduce VPD and in turn Jw, and other trees were located in a walk-in chamber where VPD and light were independently managed. R was tightly sensitive to diurnal oscillation of Jw and at negligible values of Jw (late afternoon and night) R increased. The bioristor was not sensitive to the VPD per se unless a light source was coupled to trigger Jw. This study preliminarily examined the suitability of bioristor to estimate the mean daily nutrients accumulation rate (Ca, K) in leaves comparing chemical and sensor-based procedures showing a good agreement between them opening new perspective towards the application of OECT sensor in precision agricultural cropping systems.

Seasonal and long-term consequences of esca grapevine disease on stem xylem integrity

Hydraulic failure has been extensively studied during drought-induced plant dieback, but its role in plant-pathogen interactions is under debate. During esca, a grapevine (Vitis vinifera) disease, symptomatic leaves are prone to irreversible hydraulic dysfunctions but little is known about the hydraulic integrity of perennial organs over the short- and long-term. We investigated the effects of esca on stem hydraulic integrity in naturally infected plants within a single season and across season(s). We coupled direct (ks) and indirect (kth) hydraulic conductivity measurements, and tylose and vascular pathogen detection with in vivo X-ray microtomography visualizations. Xylem occlusions (tyloses) and subsequent loss of stem hydraulic conductivity (ks) occurred in all shoots with severe symptoms (apoplexy) and in more than 60% of shoots with moderate symptoms (tiger-stripe), with no tyloses in asymptomatic shoots. In vivo stem observations demonstrated that tyloses occurred only when leaf symptoms appeared, and resulted in more than 50% loss of hydraulic conductance in 40% of symptomatic stems, unrelated to symptom age. The impact of esca on xylem integrity was only seasonal, with no long-term impact of disease history. Our study demonstrated how and to what extent a vascular disease such as esca, affecting xylem integrity, could amplify plant mortality through hydraulic failure.

BII-Implementation: The causes and consequences of plant biodiversity across scales in a rapidly changing world

The proposed Biology Integration Institute will bring together two major research institutions in the Upper Midwest—the University of Minnesota (UMN) and University of Wisconsin-Madison (UW)—to investigate the causes and consequences of plant biodiversity across scales in a rapidly changing world—from genes and molecules within cells and tissues to communities, ecosystems, landscapes and the biosphere. The Institute focuses on plant biodiversity, defined broadly to encompass the heterogeneity within life that occurs from the smallest to the largest biological scales. A premise of the Institute is that life is envisioned as occurring at different scales nested within several contrasting conceptions of biological hierarchies, defined by the separate but related fields of physiology, evolutionary biology and ecology. The Institute will emphasize the use of ‘spectral biology’—detection of biological properties based on the interaction of light energy with matter—and process-oriented predictive models to investigate the processes by which biological components at one scale give rise to emergent properties at higher scales. Through an iterative process that harnesses cutting edge technologies to observe a suite of carefully designed empirical systems—including the National Ecological Observatory Network (NEON) and some of the world’s longest running and state-of-the-art global change experiments—the Institute will advance biological understanding and theory of the causes and consequences of changes in biodiversity and at the interface of plant physiology, ecology and evolution.

INTELLECTUAL MERIT

The Institute brings together a diverse, gender-balanced and highly productive team with significant leadership experience that spans biological disciplines and career stages and is poised to integrate biology in new ways. Together, the team will harness the potential of spectral biology, experiments, observations and synthetic modeling in a manner never before possible to transform understanding of how variation within and among biological scales drives plant and ecosystem responses to global change over diurnal, seasonal and millennial time scales. In doing so, it will use and advance state-of-the-art theory. The institute team posits that the designed projects will unearth transformative understanding and biological rules at each of the various scales that will enable an unprecedented capacity to discern the linkages between physiological, ecological and evolutionary processes in relation to the multi-dimensional nature of biodiversity in this time of massive planetary change. A strength of the proposed Institute is that it leverages prior federal investments in research and formalizes partnerships with foreign institutions heavily invested in related biodiversity research. Most of the planned projects leverage existing research initiatives, infrastructure, working groups, experiments, training programs, and public outreach infrastructure, all of which are already highly synergistic and collaborative, and will bring together members of the overall research and training team.

BROADER IMPACTS

A central goal of the proposed Institute is to train the next generation of diverse integrative biologists. Post-doctoral, graduate student and undergraduate trainees, recruited from non-traditional and underrepresented groups, including through formal engagement with Native American communities, will receive a range of mentoring and training opportunities. Annual summer training workshops will be offered at UMN and UW as well as training experiences with the Global Change and Biodiversity Research Priority Program (URPP-GCB) at the University of Zurich (UZH) and through the Canadian Airborne Biodiversity Observatory (CABO). The Institute will engage diverse K-12 audiences, the general public and Native American communities through Market Science modules, Minute Earth videos, a museum exhibit and public engagement and educational activities through the Bell Museum of Natural History, the Cedar Creek Ecosystem Science Reserve (CCESR) and the Wisconsin Tribal Conservation Association.

Detection of Black Spot of Rose Based on Hyperspectral Imaging and Convolutional Neural Network

Black spot is one of the seriously damaging plant diseases in China, especially in rose production. Hyperspectral technology reflects both external features and internal structure information of measured samples, which can be used to identify the disease. In this research, both the spectral and image features of two infected roses with black spot were used to train a convolutional neural network (CNN) model. Multiple scattering correction (MSC) and standard normal variable (SNV) methods were applied to preprocess the spectral data. Cropping, median filtering and binarization were pretreatments used on the hyperspectral images. Three CNN models based on Alexnet, VGG16 and neural discriminative dimensionality reduction (NDDR) were evaluated by analyzing the classification accuracy and loss function. The results show that the CNN model based on the fusion of features has higher accuracy. The highest accuracies of detection of blackspot in different roses are 12–26 (100%) and 13–54 (99.95%), applying the NDDR-CNN model. Therefore, this research indicates that the spectral analysis based on CNN can detect black spot of roses, which provides a reference for the detection of other plant diseases, and has favorable research significance as well as prospect for development.

Remote spectral detection of biodiversity effects on forest biomass

Quantifying how biodiversity affects ecosystem functions through time over large spatial extents is needed for meeting global biodiversity goals yet is infeasible with field-based approaches alone. Imaging spectroscopy is a tool with potential to help address this challenge. Here, we demonstrate a spectral approach to assess biodiversity effects in young forests that provides insight into its underlying drivers. Using airborne imaging of a tree-diversity experiment, spectral differences among stands enabled us to quantify net biodiversity effects on stem biomass and canopy nitrogen. By subsequently partitioning these effects, we reveal how distinct processes contribute to diversity-induced differences in stand-level spectra, chemistry and biomass. Across stands, biomass overyielding was best explained by species with greater leaf nitrogen dominating upper canopies in mixtures, rather than intraspecific shifts in canopy structure or chemistry. Remote imaging spectroscopy may help to detect the form and drivers of biodiversity-ecosystem function relationships across space and time, advancing the capacity to monitor and manage Earth's ecosystems.

Machine Learning-Based Presymptomatic Detection of Rice Sheath Blight Using Spectral Profiles

Early detection of plant diseases, prior to symptom development, can allow for targeted and more proactive disease management. The objective of this study was to evaluate the use of near-infrared (NIR) spectroscopy combined with machine learning for early detection of rice sheath blight (ShB), caused by the fungus Rhizoctonia solani. We collected NIR spectra from leaves of ShB-susceptible rice (Oryza sativa L.) cultivar, Lemont, growing in a growth chamber one day following inoculation with R. solani, and prior to the development of any disease symptoms. Support vector machine (SVM) and random forest, two machine learning algorithms, were used to build and evaluate the accuracy of supervised classification-based disease predictive models. Sparse partial least squares discriminant analysis was used to confirm the results. The most accurate model comparing mock-inoculated and inoculated plants was SVM-based and had an overall testing accuracy of 86.1% (N = 72), while when control, mock-inoculated, and inoculated plants were compared the most accurate SVM model had an overall testing accuracy of 73.3% (N = 105). These results suggest that machine learning models could be developed into tools to diagnose infected but asymptomatic plants based on spectral profiles at the early stages of disease development. While testing and validation in field trials are still needed, this technique holds promise for application in the field for disease diagnosis and management.

A multi-omics approach to solving problems in plant disease ecology

The swift rise of omics-approaches allows for investigating microbial diversity and plant-microbe interactions across diverse ecological communities and spatio-temporal scales. The environment, however, is rapidly changing. The introduction of invasive species and the effects of climate change have particular impact on emerging plant diseases and managing current epidemics. It is critical, therefore, to take a holistic approach to understand how and why pathogenesis occurs in order to effectively manage for diseases given the synergies of changing environmental conditions. A multi-omics approach allows for a detailed picture of plant-microbial interactions and can ultimately allow us to build predictive models for how microbes and plants will respond to stress under environmental change. This article is designed as a primer for those interested in integrating -omic approaches into their plant disease research. We review -omics technologies salient to pathology including metabolomics, genomics, metagenomics, volatilomics, and spectranomics, and present cases where multi-omics have been successfully used for plant disease ecology. We then discuss additional limitations and pitfalls to be wary of prior to conducting an integrated research project as well as provide information about promising future directions.

Examining the Utility of Visible Near-Infrared and Optical Remote Sensing for the Early Detection of Rapid ‘Ōhi‘a Death

The early detection of plant pathogens at the landscape scale holds great promise for better managing forest ecosystem threats. In Hawai‘i, two recently described fungal species are responsible for increasingly widespread mortality in ‘ōhi‘a Metrosideros polymorpha, a foundational tree species in Hawaiian native forests. In this study, we share work from repeat laboratory and field measurements to determine if visible near-infrared and optical remote sensing can detect pre-symptomatic trees infected with these pathogens. After generating a dense time series of laboratory spectral reflectance data and red green blue (RGB) images for inoculated ‘ōhi‘a seedlings, seedlings subjected to extreme drought, and control plants, we found few obvious spectral indicators that could be used for reliable pre-symptomatic detection in the inoculated seedlings, which quickly experienced complete and total wilting following stress onset. In the field, we found similar results when we collected repeat multispectral and RGB imagery over inoculated mature trees (sudden onset of symptoms with little advance warning). We found selected vegetation indices to be reliable indicators for detecting non-specific stress in ‘ōhi‘a trees, but never providing more than five days prior warning relative to visual detection in the laboratory trials. Finally, we generated a sequence of linear support vector machine classification models from the laboratory data at time steps ranging from pre-treatment to late-stage stress. Overall classification accuracies increased with stress stage maturity, but poor model performance prior to stress onset and the sudden onset of symptoms in infected trees suggest that early detection of rapid ‘ōhi‘a death over timescales helpful for land managers remains a challenge.