2014 ASHS Annual Conference

Brazilian peach cultivar ‘Aurora-1’ requires low chilling period and its yellow, low acid flesh is adequate for fresh consumption. As the soluble solids content (SSC) and firmness have direct implications towards the consumer acceptance the determination of these two parameters is extremely important for the quality of the final product to be offered. The standard methods to quantify SSC and firmness are destructive, time consuming and sometimes require specialized procedures. Therefore non-invasive and/or non-destructive techniques have been used to determine quality parameter of fruit, e.g. the near infrared spectroscopy (NIR). Thus, the objectives of this study was to develop a methodology using NIR spectroscopy for the quality control of intact ‘Aurora-1’ peach fruit for SSC and firmness, and verify the influence of maturity stage and harvest season on the models to be developed (robustness). The models were developed with fruit obtained in 2013 at 3 maturity stages and at 3 different orchards. The spectra were collected on the background and blush color peel areas of the fruit. Spectra were obtained as log 1/R and they were not pre-treated. Principal component Analysis (PCA) was used to verify group formation and calibration was developed using Partial Least Squares (PLS) regression. Model performance was evaluated based on the values of root mean square error for prediction (RMSEP) and coefficient of determination (R²) obtained from different validation fruit samples, as such: i. full cross validation method, ii. test set of an independent dataset. PCA could not group the fruit based on maturity stages, harvest season, and background and blush peel color. In this regard the spectra obtained on the background and blush peel color were averaged and used to construct the PLS models. The model constructed using full cross validation method lead to a RMSEP of 1.05 °Brix with 11 latent variables (LVs) and a R² of 0,59. Independent dataset resulted in a less accurate model (RMSEP 1.13 °Brix, R² 0.50, 10 LV) compared to full cross validation dataset. The same trend happened for firmness determination as full cross validation resulted in better models (RMSEP 0.95 kgf, R² 0.65, 11 LV) then test set dataset (RMSEP 0.89 kgf, R² 0.05, 9 LV). Although NIR spectroscopy can be successfully used to predict SSC and firmness of intact ‘Aurora-1’ peach fruit as a non-destructive method, more chemometrics is necessary to reduce RMSEP and improve calibration model prediction accuracy.