Articles
Non-destructive techniques to determine optimal harvesting time in mango fruit
Article number
1415_17
Pages
155 – 162
Language
English
Abstract
When exported to long distance markets, mango fruit is harvested at green ripe stage.
Thus, fruit quality and flavor not always reach the market according to consumer’s requirement.
In Mexico, pulp colour and total soluble solids content are used as harvest criteria; however, both are destructive.
Some countries use the dry matter (DM) content, which is a non-destructive technique, but it is not applied because there are no local standards.
Therefore, the objectives were to build and validate a model to predict the DM content of main mango exporting varieties in Mexico.
During 2019, five groups of 40 fruits each (from unripe to already colourful) were harvested from ‘Tommy Atkins’, ‘Ataulfo’, ‘Kent’ and ‘Keitt’ cultivars.
The 200 fruits were scanned in both cheeks with an F-750® spectrometer to obtain the DM values.
The reference values were attained by a forced air oven at 60°C for 72 h.
The model was built by means of the artificial neural network application and validated during 2020 and 2021 in commercial orchards.
Dry matter was estimated applying the model generated in 2019 and loaded into an F-751® spectrometer.
The model was promising since it presented an R2=0.8487 and an RMSE=0.9059. Regarding validation, the average values of cultivars and ripening stages at harvest were 15.6% for the spectrometer and 15.2% for the conventional method, only 0.4 points of difference.
It indicates the feasibility of the F-751® to determine non-destructively the DM content of any of the cultivars; although differences were detected among them. ‘Ataulfo’ showed the best fit of the prediction model, while ‘Keitt’ had the worst.
Thus, fruit quality and flavor not always reach the market according to consumer’s requirement.
In Mexico, pulp colour and total soluble solids content are used as harvest criteria; however, both are destructive.
Some countries use the dry matter (DM) content, which is a non-destructive technique, but it is not applied because there are no local standards.
Therefore, the objectives were to build and validate a model to predict the DM content of main mango exporting varieties in Mexico.
During 2019, five groups of 40 fruits each (from unripe to already colourful) were harvested from ‘Tommy Atkins’, ‘Ataulfo’, ‘Kent’ and ‘Keitt’ cultivars.
The 200 fruits were scanned in both cheeks with an F-750® spectrometer to obtain the DM values.
The reference values were attained by a forced air oven at 60°C for 72 h.
The model was built by means of the artificial neural network application and validated during 2020 and 2021 in commercial orchards.
Dry matter was estimated applying the model generated in 2019 and loaded into an F-751® spectrometer.
The model was promising since it presented an R2=0.8487 and an RMSE=0.9059. Regarding validation, the average values of cultivars and ripening stages at harvest were 15.6% for the spectrometer and 15.2% for the conventional method, only 0.4 points of difference.
It indicates the feasibility of the F-751® to determine non-destructively the DM content of any of the cultivars; although differences were detected among them. ‘Ataulfo’ showed the best fit of the prediction model, while ‘Keitt’ had the worst.
Publication
Authors
J.A. Osuna-García, M.J. Graciano-Cristóbal, R. Goenaga
Keywords
cultivars, dry matter, spectrometer F-751
Groups involved
Online Articles (38)