BACKGROUND: Soluble solids concentration (SSC), dry matter concentration (DMC) and flesh firmness (FF) are important fruit quality parameters in stone fruits. This study investigated the ability of a commercial visible/near-infrared (NIR) spectrometer to determine SSC, DMC and FF in nectarine, peach, apricot and Japanese plum cultivars at harvest. The work was conducted in summer 2019/2020 on 14 stone fruit cultivars at Tatura, Australia. Two sub-samples of 100 fruit each were collected before and after commercial maturity (± 5 days) in order to maximize sample variability. RESULTS: Partial least square (PLS) regression models based on the second derivative of the absorbance in the 729-975 nm spectral region proved accurate for the prediction of SSC and DMC (R2 CV > 0.750). Only the model generated for SSC in 'Golden May' apricot was less precise compared to other cultivars. No visible/NIR models were accurate enough to predict FF in the cultivars under study (R2 CV < 0.750). CONCLUSION: This study demonstrated that the visible/NIR spectrometer was a reliable tool to monitor SSC and DMC in stone fruits at harvest but proved less useful for FF estimation. These results highlight the potential of visible/NIR spectrometry to evaluate stone fruit quality both in situ pre-harvest and in the laboratory after harvest.
BACKGROUND: Soluble solids concentration (SSC), dry matter concentration (DMC) and flesh firmness (FF) are important fruit quality parameters in stone fruits. This study investigated the ability of a commercial visible/near-infrared (NIR) spectrometer to determine SSC, DMC and FF in nectarine, peach, apricot and Japanese plum cultivars at harvest. The work was conducted in summer 2019/2020 on 14 stone fruit cultivars at Tatura, Australia. Two sub-samples of 100 fruit each were collected before and after commercial maturity (± 5 days) in order to maximize sample variability. RESULTS: Partial least square (PLS) regression models based on the second derivative of the absorbance in the 729-975 nm spectral region proved accurate for the prediction of SSC and DMC (R2 CV > 0.750). Only the model generated for SSC in 'Golden May' apricot was less precise compared to other cultivars. No visible/NIR models were accurate enough to predict FF in the cultivars under study (R2 CV < 0.750). CONCLUSION: This study demonstrated that the visible/NIR spectrometer was a reliable tool to monitor SSC and DMC in stone fruits at harvest but proved less useful for FF estimation. These results highlight the potential of visible/NIR spectrometry to evaluate stone fruit quality both in situ pre-harvest and in the laboratory after harvest.