President’s Faculty Research and Development Grant (PFRDG) to Test Field-applicable E-senses Methods to Detect Lethal Bronzing Disease in Sabal Palmetto Trees

January 18, 2023

Jose Ramos, Ph.D. is a Professor in NSU’s College of Computing and Engineering. He received a President’s Faculty Research and Development Grant (PFRDG) for his project titled “E-senses for Monitoring the Health of Palms with Lethal Bronzing Disease.”

Tell me about your recent grant-funded project.

This project is about detecting the lethal bronzing disease (LBD) in Sabal palmetto trees using an inexpensive electronic devices such as an electronic nose (eNose), electronic eyes (eEyes), and electronic tongues (eTongue). The reason for using such devices is that other more accurate device, such as qPCR and gas chromatography coupled to mass spectrometry (GC-MS) are not applicable in the field and require highly-trained personnel.

eNoses can detect the volatile organic compounds (VOCs) used in communication between plants and recognize when the infected palm trees release specific VOCs. eTongues work with electrochemistry and can detect flavors, much like the human tongue can. LBD is a disease caused by a phytoplasma living in the part of the plant where sap is transported (phloem tissue) and transmitted to plants by piercing-sucking insects that feed on the sap. Thus, testing the sap of Sabal Palmetto trees with an eTongue was of interest in this project. An eEyes device produces a spectrum profile as a function of wavelength, which becomes a signature that is representative of the health condition of the palm tree.

In essence, all three devices use specific sensors to capture data relevant to the health condition of the palm tree. Samples of healthy, early-infected, moderately infected, and late-stage infected palm leaves were first verified by qPCR analysis and then analyzed by the eNose, eTongue, and eEyes. The results were analyzed with Principal Component Analysis (PCA), followed by supervised machine learning to obtain a predictive model. Using a Random Forrest model, the prediction accuracy of the model was 100%.

Who is working with you on this project, and what are their roles?

I have a background in sensors and electronics, mathematical modeling, data analysis, and machine learning. As PI I test the samples with eNose, eEyes, and eTongue, and do all the data analysis, leading to the final results.

I also have multiple collaborators:

Co-PI Dr. Venkatesh Shanbhag in the Chemistry Department in the NSU Halmos College of Arts & Sciences has a background in analytical instruments and manages a garden with medicinal and ornamental plants at NSU. He assists with the use of the spectrometry equipment and designing the experiments.
Co-PI Dr. Brian Bahder in the Department of Entomology and Nematology at University of Florida-Fort Lauderdale’s Research and Education Center has a background in entomology and phytopathology with expertise in insect vectors of pathogens in tropical plants, qPCR analysis, and plant diseases, particularly the lethal bronzing disease. He conducts all the qPCR analyses, provided all the samples used in the study, directs the field investigation, and provides a chem lab structure for GC-MS analyses.
NSU biomedical engineering students Angie Castillo and Izza Kahn—Angie is instrumental in collecting data during the eEyes and eTongue experiments. Izza conducts eNose, eTongue, and eEyes experiments on specific chemical compounds found in samples of palm trees.
Postdoctoral fellow Dr. Jordana Alves Ferreira from the Brazilian Agricultural Research Corporation (Embrapa) has a Ph.D. in analytical chemistry and has studied the translocation of agricultural products as pesticides using vegetative endotherapy in coconut palm trees and has been studying the control of LBD in S. Palmetto trees in Florida. She conducts and interprets all GC-MS experiments, as well as performs qPCR analyses under the supervision of Bahder’s laboratory. She also develops specific endotherapeutic equipment for palm trees, in addition to new formulations, which were initially tested on infected palm trees, thus showing significant promise in controlling the disease.

How does this grant intersect with your work at NSU?

I teach EENG 2710 – Electric Circuits/Lab; GENG 3050 – Sensors, Measurements, and Controls; and GENG 2050 – Computer Applications in Engineering. All three courses were instrumental in carrying out this PFRDG project. On the one hand I had to build some of the devices that involve electronics and sensors, thus we used concepts from EENG 2710 and GENG 3050, and on the other hand, I had to do advanced data analysis and computing, where the concepts from GENG 2050 related.

What advice do you have for other grant seekers at NSU?

This has been a great experience, from the writing of the grant to the research part. What I like about the PFRDG program is that the format is similar to what an external grant proposal looks like. It is quite thorough and prepares you for writing an external grant. My advice is to start with a solid idea, keeping the goals realistic.

You’ll be surprised as to how many new ideas you’re going to generate during the execution of the proposal. While doing so think big—generate ideas towards extending the research to a larger grant.

My final advice is to try to team up with people who are going to help you take your research to a higher level. In my case I benefitted immensely from my Co-PI’s expertise in the field and him being so well known, along with his lab facilities at UF-IFAS, and his postdoc, Dr. Jordana Ferreira, with whom I interacted a great deal.

What is the next step in this project?

This project has given us many insights in different directions. We were able to prove that we could detect LBD using eNose, eEyes, and eTongue. The literature has shown many articles where the eNose has been used for detecting different kinds of cancers, gastrointestinal diseases, etc. However, the use of all three instruments in detection problems has never been done before. In that sense, this PFRDG project was a pioneering effort. We would like to try all three instruments in the medical field, but this requires partnering with a team that has all the protocols in place.

What is the next grant proposal or project on your agenda?

From this PFRDG grant we learned that the citrus industry is going through a substantial crisis with the so-called “Citrus Greening” disease. We did a thorough search about this disease and found out that the disease dynamics are similar to the LBD: phytoplasma, which is transmitted by vector insects. Although, for citrus trees, the insect and the phytoplasma are of different classes and species. Thus, I contacted the USDA and had a Zoom meeting with three grant officers. They all agreed that combining our research with a citrus direction would increase our funding chances significantly. So, our next step is to write a grant to try to combine our current research on LBD with the detection of Citrus Greening disease using sensors and field-based technology.