Near infrared reflectance spectroscopy to quantify Perkinsus marinus infecting Crassostrea virginica

Abstract Disease caused by the protozoan parasite Perkinsus marinus remains a challenge for fisheries and aquaculture of the eastern oyster, Crassostrea virginica. Near infrared reflectance (NIR) spectroscopy is a technology that allows measurements of multiple biochemical components at the same time, providing extensive information for every tested sample using a single, high throughput, rapid assay. Adding another parameter such as determination of P. marinus level of infection may be of great interest to assess, along with other biochemical parameters, oyster's general condition and also characterize oysters that present resistance or tolerance to this problematic parasite. During this study, the NIR spectroscopy technology was used to test the feasibility of an alternative means of quantitative detection of P. marinus in the eastern oyster. In order to accomplish this, two experiments were performed: a P. marinus “spike” experiment and a natural infection study. During the spike experiment, pathogen-free, homogenized C. virginica slurries were spiked with known concentrations of cultured P. marinus cells. NIR spectroscopy spectra were collected on these preparations as is (wet) and after freeze drying (dry). During the natural infection study, oysters from the upper portion of the Rappahannock River were transferred to the York River, where P. marinus is endemic. Subsampled monthly for 7 months and diagnosed for the presence and intensity of acquired infections. NIR spectroscopy spectra were collected on freeze dried samples and P. marinus concentration was determined using quantitative PCR (qPCR). Calibration models were developed using modified partial least squares regression (PLS) based on cross-validation and tested using a validation set. Comparison of reference values to predicted values, showed a high correlation for all models: for the spike experiment reaching a R2val of 0.70 and 0.93 for the wet and dry model respectively; and for the natural infection model reaching a R2val of 0.81 for the dry model. These metrics indicate that P. marinus cells present a quantifiable spectral signature against the background of C. virginica oyster tissues and implies that NIR spectroscopy is a promising technology to detect protozoan infection in animal tissues. This technology seems unlikely to replace qPCR or other classical methods like RFTM, primarily because of the amount of preparation needed to effect NIR sampling. However, NIR spectroscopy is used for assessing other facets of oyster condition simultaneously, such as glycogen and lipid, and may be used to obtain reasonable estimates of P. marinus infection when more traditional methods could not or were not used.