Skip to main content
ARS Home » Southeast Area » Griffin, Georgia » Plant Genetic Resources Conservation Unit » Research » Research Project #446521

Research Project: Characterizing the Solanum Melongena Germplasm Collection for Harvest-Mature Fruit Shape, Size, and Color Using Detached Fruit Images

Location: Plant Genetic Resources Conservation Unit

Project Number: 6046-21000-013-042-S
Project Type: Non-Assistance Cooperative Agreement

Start Date: Aug 1, 2024
End Date: Jul 31, 2025

Objective:
Eggplant varieties vary significantly at harvest maturity in terms of color (including green, white, and light- and dark-purple fruited types); size (ranging from 3 to 12 cm in diameter); and shape (with a length-to-width ratio ranging from <1 to >5). Fruit with different morphological characteristics represent distinct market classes with varying popularities in different regions of the world. For example, varieties with long, purple fruit are popular in China; small, round, green-white fruit in Thailand; and small, round, dark-purple fruit in India. While historically United States production has predominantly consisted of large-fruited, purple, teardrop-shaped types, other varieties appealing to different ethnic communities represent expanding market opportunities for farmers. This proposal seeks to characterize the eggplant collection for harvest-mature (commercial picking stage) fruit shape, size, and color in order to promote its use in breeding for diverse market classes. We plan to collect quantitative morphological data collected via automated image analysis from photographs of harvested fruit, to complement the qualitative morphological fruit shape ratings that already exist in GRIN for the eggplant collection. Furthermore, this proposal would significantly expand the number of S. melongena accessions in GRIN that feature fruit photos - there are currently only 54, some of which display fruit at their physiological, rather than harvest maturity. There is no crop germplasm committee (CGC) for eggplant, and therefore no crop vulnerability statement available. However, in the most recent vulnerability statement for pepper, characterization of mature fruit color was ranked seventh out of forty traits, demonstrating the value of fruit morphological data to the breeding and genetics community for a very similar crop.

Approach:
Fruit will be picked at harvest maturity, defined as the largest fruit size attained per accession before any color or firmness changes associated with ripening begin. Because this will vary based on the maturity of each accession, we will rate plots for maturity and pick fruit up to three times a week beginning likely in mid-July. Fruit will be bulked across the five plants in a plot and transported to the Cooperator's lab. Up to 10 fruit per accession will be placed on a blue background in a light box with uniform lighting, along with a color card and a QR-encoded label, and imaged with a digital single-lens reflex camera. Custom image analysis scripts using the python open computer vision library will be used to extract multiple features from each fruit in each image: length, width, length-width ratio, area, fruit color in LAB color space, calyx color in LAB color space, roundness, and solidity (how well the fruit shape fits inside a best-fit ellipsoid). Both means and variances for each trait will be calculated for each accession. Image photos and extracted shape, color, and size statistics to GRIN.