Prediction of blueberry sensory texture attributes by integrating multiple instrumental measurements

Authors: OH, HEEDUK - North Carolina State University; STAPLETON, LEE - Sensory Spectrum, Inc; GIONGO, LARA - Fondazione Edmund Mach; Johanningsmeier, Suzanne; MOLLINARI, MARCELO - North Carolina State University; MAINLAND, CHARLES - North Carolina State University; PERKINS-VEAZIE, PENELOPE - North Carolina State University; IORIZZO, MASSIMO - North Carolina State University

Submitted to: Postharvest Biology and Technology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/18/2024
Publication Date: 8/28/2024
Citation: Oh, H., Stapleton, L., Giongo, L., Johanningsmeier, S.D., Mollinari, M., Mainland, C., Perkins-Veazie, P., Iorizzo, M. 2024. Prediction of blueberry sensory texture attributes by integrating multiple instrumental measurements. Postharvest Biology and Technology. 218. Article 113160. https://doi.org/10.1016/j.postharvbio.2024.113160.
DOI: https://doi.org/10.1016/j.postharvbio.2024.113160

Interpretive Summary: Blueberry (Vaccinium spp.) texture contributes to consumer satisfaction, shelf-life, and machine harvestability and is now a crucial goal in breeding for new fresh market cultivars. This study characterized the fruit texture profile of 43 blueberry cultivars and measured their mechanical properties by two commonly used instrumental methods. The main differences in blueberry textures were springiness, hardness, and snap/crisp attributes. These sensory attributes could be reasonably estimated with single parameters from common instrumental texture measurements, such as penetration and compression tests. Prediction of more complex textures, such as juiciness, was possible through development of a multivariate model that used 17 parameters extracted from the data generated by instrumental texture profile analysis. This study provides a basis for blueberry breeding programs to utilize additional instrumental parameters to improve estimations of sensorial textures for enhanced selection of cultivars with desired textures.

Technical Abstract: Blueberry (Vaccinium spp.) texture contributes to consumer satisfaction, shelf-life, and machine harvestability and is now a critical goal in breeding for new fresh market cultivars. The industry commonly uses instrumental methods to phenotype texture, assuming that instrumental measurements correlate with sensory perceptions. However, the relationship between perceived sensory textures and mechanical parameters is not well established. In this study, we characterized the fruit texture profile of 43 blueberry cultivars using nine sensory descriptors and determined the predictability of the sensory attributes using mechanical parameters. The sensory study was done by a trained descriptive sensory analysis panel, and instrumental analysis was performed using flat probe penetration and texture profile analysis (TPA) methods. Differences in the perceived firmness of blueberries were mainly due to springiness, hardness, and snap/crisp attributes. Among the mechanical parameters, maximum force (FM; a flat probe penetration parameter) and gumminess (a TPA parameter) had the strongest correlations with these three sensory attributes. To develop predictive models for the nine sensory attributes, multivariate statistical methods were used. The highest level of prediction accuracy was achieved when all the penetration or TPA parameters were used for model development. The R^2 values increased by up to 0.66 compared to using a single mechanical parameter. Springiness, hardness, and snap/crisp were predictable with R2 > 0.5, regardless of the instrumental method used. TPA parameters were more suitable for predicting juiciness while residual skin was only predictable using penetration parameters. Mealiness was not predictable with any instrumental measure (R^2 < 0.05). For most of the sensory attributes, the models were able to effectively discern the cultivars with the highest or lowest intensity scores. This study provides a basis for breeding programs to utilize improved estimations of the sensorial texture using diverse mechanical parameters to enhance selection for desired blueberry textures.