Macronutrient deficiency reduces growth and influences vegetation indices of greenhouse grown ornamental and vegetable plants as measured by the TraitFinder digital phenotyping system.

Introduction: Recent technological advances in high resolution image capture and analysis have led to increased adoption of high-throughput digital phenotyping in plant science research. High-throughput digital phenotyping provides a nondestructive method to quantify changes in plant growth and health in response to environmental factors or developmental cues. Moreover, it allows researchers to conduct large experiments in a time- and cost-efficient manner. The TraitFinder is a digital phenotyping system developed by Phenospex (Heerlen, Netherlands) that measures plant morphological (e.g., digital biomass) and spectral (leaf light reflectance) information. Leaf reflectance is presented as five vegetation indices (e.g., normalized difference vegetation index). Methods: This project evaluated nitrogen (N), phosphorus (P), and potassium (K) deficiency in greenhouse grown ornamental and vegetable plants using the TraitFinder. Plant species included celosia, coleus, marigold, petunia, and tomato. Plants were fertilized with a complete Hoagland's solution (control), and three modified solutions: Hoagland's solution without nitrogen (-N), phosphorus (-P), or potassium (-K). Each plant species was evaluated separately with eight replicate plants per treatment, organized as a randomized complete block design. Results: Treatment with -N, -P, and -K solutions resulted in reduced vegetative growth and decreased concentration of the corresponding macronutrient in leaf tissue for all species evaluated. We observed that the presence of flowers would negatively affect calculations of the vegetation indices due to their distinct spectral properties; therefore, flowers must be excluded to accurately quantify plant health parameters. In general, we observed a common trend where GLI (green leaf index) and NDVI (normalized difference vegetation index) decreased, and NPCI (normalized pigment chlorophyll index) and PSRI (plant senescence reflectance index) increased in response to macronutrient deficiency. The measure of GLI, NDVI, NPCI, and PSRI were different from the control plants, but these observations were dependent on the nutrient deficiency and species tested. Discussion: Our results underscore the importance of accounting for species-specific spectral signatures when assessing plant responses to nutrient deficiencies. This project also provides reference values for interpreting vegetation indices, offering valuable guidance for scientists implementing digital phenotyping in their experimental protocols. Digital phenotyping can significantly improve experimental throughput and provide quantitative insights into plant health.