PATHOGEN DETECTION AND MONITORING FROM INDUCED VOCS OF CITRUS VARIETALS

High value fruit and nut crops are a vital component of agricultural productivity in the United States. Almost 80% of this economic sector is located in the Pacific Northwest (Washington and Oregon), California, and the Southeast (Texas and Florida). Orchards and vineyards in these regions provide healthful fruits and nuts to the nation and for export, support local tourism, and generate economic and environmental benefits. However, orchard and vineyard plants are vulnerable to numerous diseases caused by pests and microbial agents, making our food supply susceptible to both naturally-occurring and intentionally-released pathogens. Huanglongbing (HLB) is a pathogen that devastates citrus crops and the citrus industry desperately needs early detection of these infections. However, there are no robust diagnostic tests currently available for early identification of this or other citrus pathogens under field conditions. This proposal directly responds to the need to reduce the impact of plant pathogens and insect pests on citrus crops by developing a means of early detection. Our proposed technology senses and measures the odor signatures that emanate from the leaves of the plant to identify HLB infected trees from healthy trees. The odor signatures are composed of volatile organic compounds (VOCs) with varying abundances that are different between healthy plants and diseased plants and from plants infected by different pathogens. Our prior work has shown that VOC signature identification can be effective at detecting HLB infected trees before the infection is visually apparent and before it can be confirmed by Polymerase Chain Reaction (PCR) analysis. The VOC signature detection technology proposed has been shown to be effective on sweet orange varietals (Hamlin, Valencia and Madame Vinous). The objectives will be to demonstrate that the VOC signature for HLB infection as detected by the EZKnowz tool is robust against background variations, evaluate the HLB VOC "fingerprint" across many citrus varietals, and optimize the EZKnowz instrument for performance optimization in urban areas and in commercial orchards. Anticipated Results and Potential Commercial Applications: The Phase I efforts will provide the information needed to build the engineering prototype instrument for economical use in commercial orchards and urban landscapes. The Phase II effort will demonstrate economic viability by developing a tool that will be sensitive to the HLB signature and sufficiently fast such that it can be useful beyond academic research. The technical benefit will be a deeper understanding of the response of plants to HLB infections across many citrus varieties. The economic benefit will be a tool that can be used in finding and managing HLB infestations in citrus trees. In Florida, the study by Hodges and Spreen estimates the difference between the scenarios with- versus without-HLB is a total output impact of −$4.541 billion, total value added of −$2.717 billion, total labor income of −$1.753 billion, and average total employment of −8,257 jobs. The impact of a similar infestation in California would likely be larger.

Task 1: Test against interfering VOC signals In this task, we will collect VOCs from HLB-infected citrus trees in urban environments and test against confounding VOCs. The health status determination of a tree is based on differences in abundances of various biogenic VOCs produced by tree foliage. If any of the extraneous chemicals overlap with the measured VOC compounds, the accuracy of diagnosis could be diminished. Therefore, it is important to account for most likely scenarios for chemical noise that can occur during field measurements, both in orchards and in urban environment. We are planning on performing a set of experiments using EZKnowz(TM) platform to record typical chemical backgrounds. In orchard setting these will include: typical pesticides/herbicides employed by citrus growers, rotting vegetation and decomposing citrus fruits, blossoms. In urban setting, the tested backgrounds will include: petroleum products, gasoline/diesel fumes, non-specific extraneous odors (e.g. asphalt, cooking fumes, outdoor pets, etc.) The GC/DMS profiles of the recorded chemicals backgrounds will be evaluated in order to determine the sources of greatest interferences and degree of overlap with the discriminating features of the GC/DMS signatures of HLB-infected citrus trees. A numerical algorithm will be developed to allow for background correction. Task 2: Test variability of HLB infection across several citrus varieties and locations In this task we will collect VOCs from HLB-infected citrus trees from as many geographic locations and varietals as possible: Southern California urban environment, multiple cultivars including grafted plants, South Texas sweet orange and grapefruit, Florida in orchards (Hamlin sweet orange), Greenhouse studies. Task 3: Algorithm development and data analysis. We plan to use algorithm approaches to perform two critical functions for this program: (1) to identify and limit the confounding effect of transient or persistent chemical environmental signatures that may hinder correct tree diagnosis; and (2) to down-select the most appropriate HLB VOC markers that occur in a wide range of commercial and ornamental citrus varietals. Task 4: Design optimization This effort is to reduce sampling and analysis time to