Epsilon Open Archive

Berries are used for producing a wide variety of products: fresh berries, beverages based on berry
juice and powders used in cosmetics and health supplements. The berry industry depends on the
collection of wild-growing berries in the forest during a growing season and there is a variation in
concentration of important substances over this growing season. Usually there is an optimal time
during the season when the berries are most rich in antioxidants. The traditional quality
measurement is done by wet chemical methods that are accurate but slow and expensive and
therefore it would be desirable to have a faster and cheaper method. In the project we compared
wet chemical methods of analysis with near infrared spectroscopy (NIR). This technique can be
used in the field and gives fast results (within 2 minutes).
A systematic study was carried out under a growing season in Finland and Sweden. Bilberries
were collected in the field with regular intervals. The berries were measured by NIR spectroscopy
and then sent for chemical analysis. The study also included storage and homogenization tests.
The important variables for describing bilberries are anthocyanin content, degrees brix and dry
solids. As for the treatments of the berries, NIR measurement on homogenized fresh berries
worked best and with whole berries the results were slightly worse. The results for frozen berries
were less useful.
The NIR data obtained were analyzed in exploratory fashion by PCA analysis. This is fast
analysis which often can show preliminary results. A trend with time could be observed. The lab data results were analyzed by ANOVA. Significant results were obtained for antioxidant
concentrations, but not for dry solids. Regression analysis between NIR spectra and lab results
showed that NIR can be used to predict certain properties of homogenized fresh berries. The
present study was a small scale study carried out during only one growing season. It worked very
well as a proof of principle study. For better models relating NIR spectra to lab results, larger
studies over more growing seasons will be needed.