Classification of Plant Electrophysiology Signals for Detection of Spider Mites Infestation in Tomatoes

Early detection of dark-affected plant mechanical responses using enhanced electrical signals

BACKGROUND

Mechanical damage to plants triggers local and systemic electrical signals that are eventually decoded into plant defense responses. These responses are constantly affected by other environmental stimuli in nature, for instance, light fluctuation. In recent years, studies on decoding plant electrical signals powered by various machine learning models are increasing in a sense of early prediction or detection of different environmental stresses that threaten plant growth or crop yields. However, the main bottleneck is the low-throughput nature of plant electrical signals, making it challenging to obtain a substantial amount of training data. Consequently, training these models with small datasets often leads to unsatisfactory performance.

RESULTS

In the present work, we set out to decode wound-induced electrical signals (also termed slow wave potentials, SWPs) from plants that are deprived of light to different extents. Using non-invasive electrophysiology, we separately collected sets of local and distal SWPs from the treated plants. Then, we proposed a workflow based on few-shot learning to automatically identify SWPs. This workflow incorporates data preprocessing, feature extraction, data augmentation and classifier training. We established the integral and the first-order derivative as features for efficiently classifying SWPs. We then proposed an Adversarial Autoencoder (AAE) structure to augment the SWP samples. Combining them, the Random Forest classifier allowed remarkable classification accuracies of 0.99 for both local and systemic SWPs. In addition, in comparison to two other reported methods, our proposed AAE structure enabled better classification results using our tested features and classifiers.

CONCLUSIONS

The results of this study establish new features for efficiently classifying wound-induced electrical signals, which allow for distinguishing dark-affected local and systemic plant wound responses. We also propose a new data augmentation structure to generate virtual plant electrical signals. The methods proposed in this study could be further applied to build models for crop plants using electrical signals as inputs, and also to process other small-scale signals.

Detecting stress caused by nitrogen deficit using deep learning techniques applied on plant electrophysiological data

Plant electrophysiology carries a strong potential for assessing the health of a plant. Current literature for the classification of plant electrophysiology generally comprises classical methods based on signal features that portray a simplification of the raw data and introduce a high computational cost. The Deep Learning (DL) techniques automatically learn the classification targets from the input data, overcoming the need for precalculated features. However, they are scarcely explored for identifying plant stress on electrophysiological recordings. This study applies DL techniques to the raw electrophysiological data from 16 tomato plants growing in typical production conditions to detect the presence of stress caused by a nitrogen deficiency. The proposed approach predicts the stressed state with an accuracy of around 88%, which could be increased to over 96% using a combination of the obtained prediction confidences. It outperforms the current state-of-the-art with over 8% higher accuracy and a potential for a direct application in production conditions. Moreover, the proposed approach demonstrates the ability to detect the presence of stress at its early stage. Overall, the presented findings suggest new means to automatize and improve agricultural practices with the aim of sustainability.

The Electrome of a Parasitic Plant in a Putative State of Attention Increases the Energy of Low Band Frequency Waves: A Comparative Study with Neural Systems

Selective attention is an important cognitive phenomenon that allows organisms to flexibly engage with certain environmental cues or activities while ignoring others, permitting optimal behaviour. It has been proposed that selective attention can be present in many different animal species and, more recently, in plants. The phenomenon of attention in plants would be reflected in its electrophysiological activity, possibly being observable through electrophytographic (EPG) techniques. Former EPG time series obtained from the parasitic plant Cuscuta racemosa in a putative state of attention towards two different potential hosts, the suitable bean (Phaseolus vulgaris) and the unsuitable wheat (Triticum aestivum), were revisited. Here, we investigated the potential existence of different band frequencies (including low, delta, theta, mu, alpha, beta, and gamma waves) using a protocol adapted from neuroscientific research. Average band power (ABP) was used to analyse the energy distribution of each band frequency in the EPG signals, and time dispersion analysis of features (TDAF) was used to explore the variations in the energy of each band. Our findings indicated that most band waves were centred in the lower frequencies. We also observed that C. racemosa invested more energy in these low-frequency waves when suitable hosts were present. However, we also noted peaks of energy investment in all the band frequencies, which may be linked to extremely low oscillatory electrical signals in the entire tissue. Overall, the presence of suitable hosts induced a higher energy power, which supports the hypothesis of attention in plants. We further discuss and compare our results with generic neural systems.

Stimulus classification with electrical potential and impedance of living plants: comparing discriminant analysis and deep-learning methods

The physiology of living organisms, such as living plants, is complex and particularly difficult to understand on a macroscopic, organism-holistic level. Among the many options for studying plant physiology, electrical potential and tissue impedance are arguably simple measurement techniques that can be used to gather plant-level information. Despite the many possible uses, our research is exclusively driven by the idea of phytosensing, that is, interpreting living plants' signals to gather information about surrounding environmental conditions. As ready-to-use plant-level physiological models are not available, we consider the plant as a blackbox and apply statistics and machine learning to automatically interpret measured signals. In simple plant experiments, we exposeZamioculcas zamiifoliaandSolanum lycopersicum(tomato) to four different stimuli: wind, heat, red light and blue light. We measure electrical potential and tissue impedance signals. Given these signals, we evaluate a large variety of methods from statistical discriminant analysis and from deep learning, for the classification problem of determining the stimulus to which the plant was exposed. We identify a set of methods that successfully classify stimuli with good accuracy, without a clear winner. The statistical approach is competitive, partially depending on data availability for the machine learning approach. Our extensive results show the feasibility of the blackbox approach and can be used in future research to select appropriate classifier techniques for a given use case. In our own future research, we will exploit these methods to derive a phytosensing approach to monitoring air pollution in urban areas.

Calcium Signaling in Plant-Insect Interactions

In plant-insect interactions, calcium (Ca²⁺) variations are among the earliest events associated with the plant perception of biotic stress. Upon herbivory, Ca²⁺ waves travel long distances to transmit and convert the local signal to a systemic defense program. Reactive oxygen species (ROS), Ca²⁺ and electrical signaling are interlinked to form a network supporting rapid signal transmission, whereas the Ca²⁺ message is decoded and relayed by Ca²⁺-binding proteins (including calmodulin, Ca²⁺-dependent protein kinases, annexins and calcineurin B-like proteins). Monitoring the generation of Ca²⁺ signals at the whole plant or cell level and their long-distance propagation during biotic interactions requires innovative imaging techniques based on sensitive sensors and using genetically encoded indicators. This review summarizes the recent advances in Ca²⁺ signaling upon herbivory and reviews the most recent Ca²⁺ imaging techniques and methods.

Classification of various nutrient deficiencies in tomato plants through electrophysiological signal decomposition and sample space reduction

Plants leave testimonies of undergoing physical state by depicting distinct variations in their electrophysiological data. Adequate nutrition of plants signifies their role in the growth and a plentiful harvest. Plant signal data carries enough information to detect and analyse nutrient deficiency. Classification of nutrient deficiencies through signal decomposition and bilevel measurements has not been reported earlier. The proposed work explores tomato plants in four-time cycles (Early Morning, Morning, After Noon, Night) of macronutrients Calcium (Ca), Nitrogen (N) and micronutrients Manganese (Mn), Iron (Fe). Using the Empirical Mode Decomposition method (EMD), signals are decomposed into Intrinsic Mode Functions (IMF) in 10-levels. Further, Intrinsic mode functions are grouped into two clusters to extract descriptive data statistics and bi-level measurements. Then a novel sample selection method is proposed to achieve a better classification rate by reducing sample space. A binary classification model is built to train and test 15 features individually using discriminant analysis and naïve-Bayes classifier variants. The reported results achieved a classification rate up to 98% after 5-fold cross-validation. Attained findings endorse novel pathways for detection and classification of nutrient deficiencies in the early stages, consequently promoting prevention and treatment approaches earliest to the appearance of symptoms, also helping to enhance plant growth.

Do plants pay attention? A possible phenomenological-empirical approach

Attention is the important ability of flexibly controlling limited cognitive resources. It ensures that organisms engage with the activities and stimuli that are relevant to their survival. Despite the cognitive capabilities of plants and their complex behavioural repertoire, the study of attention in plants has been largely neglected. In this article, we advance the hypothesis that plants are endowed with the ability of attaining attentive states. We depart from a transdisciplinary basis of philosophy, psychology, physics and plant ecophysiology to propose a framework that seeks to explain how plant attention might operate and how it could be studied empirically. In particular, the phenomenological approach seems particularly important to explain plant attention theoretically, and plant electrophysiology seems particularly suited to study it empirically. We propose the use of electrophysiological techniques as a viable way for studying it, and we revisit previous work to support our hypothesis. We conclude this essay with some remarks on future directions for the study of plant attention and its implications to botany.

Rapid systemic responses to herbivory

Rapid systemic signals travel within the first seconds and minutes after herbivore infestation to mount defense responses in distal tissues. Recent studies have revealed that wound-induced hydraulic pressure changes play an important role in systemic electrical signaling and subsequent calcium and reactive oxygen species waves. These insights raise new questions about signal specificity, the role of insect feeding guild and feeding style and the impact on longer term plant defenses. Here, we integrate the current molecular understanding of wound-induced rapid systemic signaling in the framework of insect-plant interactions.

Field-Effect Transistor-Based Biosensors for Environmental and Agricultural Monitoring

The precise monitoring of environmental contaminants and agricultural plant stress factors, respectively responsible for damages to our ecosystems and crop losses, has nowadays become a topic of uttermost importance. This is also highlighted by the recent introduction of the so-called "Sustainable Development Goals" of the United Nations, which aim at reducing pollutants while implementing more sustainable food production practices, leading to a reduced impact on all ecosystems. In this context, the standard methods currently used in these fields represent a sub-optimal solution, being expensive, laboratory-based techniques, and typically requiring trained personnel with high expertise. Recent advances in both biotechnology and material science have led to the emergence of new sensing (and biosensing) technologies, enabling low-cost, precise, and real-time detection. An especially interesting category of biosensors is represented by field-effect transistor-based biosensors (bio-FETs), which enable the possibility of performing in situ, continuous, selective, and sensitive measurements of a wide palette of different parameters of interest. Furthermore, bio-FETs offer the possibility of being fabricated using innovative and sustainable materials, employing various device configurations, each customized for a specific application. In the specific field of environmental and agricultural monitoring, the exploitation of these devices is particularly attractive as it paves the way to early detection and intervention strategies useful to limit, or even completely avoid negative outcomes (such as diseases to animals or ecosystems losses). This review focuses exactly on bio-FETs for environmental and agricultural monitoring, highlighting the recent and most relevant studies. First, bio-FET technology is introduced, followed by a detailed description of the the most commonly employed configurations, the available device fabrication techniques, as well as the specific materials and recognition elements. Then, examples of studies employing bio-FETs for environmental and agricultural monitoring are presented, highlighting in detail advantages and disadvantages of available examples. Finally, in the discussion, the major challenges to be overcome (e.g., short device lifetime, small sensitivity and selectivity in complex media) are critically presented. Despite the current limitations and challenges, this review clearly shows that bio-FETs are extremely promising for new and disruptive innovations in these areas and others.

Identifying General Stress in Commercial Tomatoes Based on Machine Learning Applied to Plant Electrophysiology

Automated monitoring of plant health is becoming a crucial component for optimizing agricultural production. Recently, several studies have shown that plant electrophysiology could be used as a tool to determine plant status related to applied stressors. However, to the best of our knowledge, there have been no studies relating electrical plant response to general stress responses as a proxy for plant health. This study models general stress of plants exposed to either biotic or abiotic stressors, namely drought, nutrient deficiencies or infestation with spider mites, using electrophysiological signals acquired from 36 plants. Moreover, in the signal processing procedure, the proposed workflow reuses information from the previous steps, therefore considerably reducing computation time regarding recent related approaches in the literature. Careful choice of the principal parameters leads to a classification of the general stress in plants with more than 80% accuracy. The main descriptive statistics measured together with the Hjorth complexity provide the most discriminative information for such classification. The presented findings open new paths to explore for improved monitoring of plant health.