RAPID PREDICTION OF BIOCHEMICAL COMPOSITION IN EASTERN OYSTER Crassostrea virginica USING NEAR INFRARED REFLECTANCE SPECTROSCOPY.  

Eric Guévélou*, Standish K. Allen, Jr.
Aquaculture Genetics and Breeding Technology Center
Virginia Institute of Marine Science
Gloucester Point, Virginia 23062. United States of America.
EGuevelou@vims.edu

Near infrared reflectance spectroscopy (NIRS) is a high-throughput, cost-effective, analytical tool used over the last decades in various sectors like pharmaceutical or food industries. The principle of NIRS is based on absorption of the energy from the infrared spectrum by specific chemical bonds. A high performance wide-band halogen lamp inside the NIRS spectrometer emits light into samples causing characteristic vibrations. The resulting reflected light (wavelength region 1000 to 2500 nm) provides a fingerprint (light spectra) on the composition of these chemical bonds. These spectra coupled with actual measurements of the parameters of interest can be modelled using software to create tools to identify and quantify specific substances, allowing the rapid and precise measurements of multiple samples.

We developed NIRS models to determine compositional analysis of moisture and glycogen in the Eastern oyster Crassostrea virginica. Oysters tissues were sampled and homogenized individually, then scanned by NIRS spectrometer. These same samples were analyzed for moisture and glycogen content using traditional methods of freeze-drying and the colorimetric iodine method, respectively. Calibration models were determined and NIRS-prediction tools were developed for both moisture and glycogen content. Validation of the models, based on the comparison of model predicted and measured data, showed high correlation (R² ≥ 0.94, ratio of performance deviation ≥ 4.2) for both parameters. The overall resulting calibration and validation parameters allowed the use of these models for quantitative applications.

To the best of our knowledge, this is the first attempt to develop NIRS calibration to predict the composition of Eastern oyster C. virginica. A specific interest for Aquaculture Genetics and Breeding Technology Center is to derive compositional data for use with our selective breeding programs.