Quality Assessment of Growing Media with Near-Infrared Spectroscopy: Chemical Characteristics and Plant Assays

• Published in: European journal of horticultural science 2008-02, Vol.73 (1), p.28-36
• Main Authors: Terhoeven-Urselmans, T., Bruns, C., Schmilewski, G., Ludwig, B.
• Format: Article
• Language: English
• Subjects: Cabbages; Growing media; Minerals; Nitrogen; Peat; Plants; Salts; Spectral reflectance; Spectroscopy; Wavelengths
• ISSN: 1611-4426; 1611-4434

Quality control of growing media mainly consists of chemical analysis and plant assays, which are time-consuming and expensive. Objectives were to test, if near-infrared spectroscopy (NIRS) is capable to predict several chemical characteristics and plant assay results for a large population of various peat-based growing media having a wide range of humification degrees and characteristics (n=320). Near-infrared measurements (including the visible range, 400–2,500 nm) were done with fresh and with dried and ground growing media in order to predict their chemical characteristics and the results of plant assays using Chinese white cabbage (Brassica napus var. Chinensis). Spectral manipulations (taking 1st to 3rd derivative after baseline correction), cross-validation and a modified partial-least squares regression method were used to develop equations over the whole spectrum. Generally, NIRS predicted the chemical characteristics of growing media and the yields of fresh weight of Chinese white cabbage and rating better for fresh than for dried and ground samples. The pH and contents of total carbon and nitrogen, salt, P, K, mineral nitrogen, NO3-, NH4+ and the NH4+:NO3- ratio were predicted well: the RSC (ratio of standard deviation of laboratory results to the standard error of cross-validation) ranged between 2.0 (NO3-) and 4.4 (total carbon), the correlation coefficient (r) of measured against predicted values was higher or equal to 0.9 and the regression coefficient (a) was between 0.9 and 1.1. The good predictions of total carbon content may have been partly due to a clustering of data. Fresh weight yield of Chinese white cabbage was predicted well for the subpopulation of the growing medium with a degree of humification of H2 to H3 on the von Post humification scale (RSC=2.0, r=0.9 and a=0.9). The fresh weight yields for the subpopulations of growing media with H5 to H6 and with H7 were predicted satisfactorily (RSC =1.6 and 1.9, respectively). The prediction of the rating at harvest (overall plant impression) was satisfactorily for two subpopulations (RSC=1.7, r=0.8 and a=0.9 or 1.0) but unsatisfactory for the one with H5 to H6 (RSC=1.3). Overall, NIRS is sufficiently reliable to be used for standard chemical analysis for growing media and it is promising in predicting the results of plant assays.