Abstract

Journal of Near Infrared Spectroscopy
Volume 17 Issue 5, Pages 289–301 (2009)
doi: 10.1255/jnirs.850

Predictive capacity of visible-near infrared spectroscopy for quality parameter assessment of compost

Sandra Herrmann,^a Jochen Mayer,^b Kerstin Michel^a,* and Bernard Ludwig^a
aDepartment of Environmental Chemistry, University of Kassel, Nordbahnhofstr. 1a, D-37213 Witzenhausen, Germany. E-mail: kmichel@uni-kassel.de
^bAgroscope Reckenholz-Tänikon Research Station ART, Reckenholzstr. 191, CH-8046 Zurich, Switzerland

Screening tests are basic procedures commonly used to assess compost quality. Important parameters for quality assessment are the germination capacity and the suppression of plant pathogens which have to be measured by time-consuming laboratory methods. The objective was to test whether visible (vis) and near infrared (NIR) spectroscopy (vis-NIR) is useful to analyse parameters important for compost quality. Ninety seven compost samples from Switzerland were analysed by conventional methods and by vis-NIR. The content of organic (C_org) and inorganic C (C_inorg), total N (N_tot), mineralisable N after 56 days (N_{min_d56}), total P (P_tot), K, Ca and salt, the C/N ratio, pH and microbiological characteristics [hydrolysis of fluorescein diacetate (FDA- hydrolysis) as indicator of total enzyme activity and cellulase activity] were determined. Furthermore, plant tolerance and the suppression of pathogens were tested using germination tests with salad, cress, ryegrass and bean or a Rhizoctonia solani bioassay, respectively. The samples were scanned in the range of 400–2500 nm (visible light and NIR) using a Foss NIRSystems spectrometer 6500. A modified partial least squares regression method and the whole spectrum were used to develop cross-validation equations for all constituents. For this, the first to third derivative was calculated. The prediction accuracy was evaluated as excellent for C_org and good for N_t and the C/N ratio based on the RSC values (ratio of standard deviation of laboratory results to standard error of cross-validation) and the coefficients of determination (r²). Approximate quantitative predictions were possible for the contents of P_tot, K, Ca and salt, whereas for the constituents C_inorg, N_{min_d56}, FDA-hydrolysis and the germination tests with cress and salad only between high and low values could be discriminated. Unsuccessful predictions as indicated by RSC values lower than 1.5 and r² values below 0.50 were obtained for pH, cellulase activity, germination tests with ryegrass and bean and the disease suppression test using R. solani. Overall the results of the present study indicate that vis-NIR spectroscopy has the potential to be used for quality assessment of composts and to replace time-consuming methods such as germination tests using salad and cress. However, the use for monitoring purposes requires further research to clarify whether other complex quality parameters such as disease suppression indicators may also be predicted successfully.

Keywords: compost, near-infrared spectroscopy, prediction, MPLS, germination test, disease suppression indicator

---

Full-text article (682 kB) (subscribers only)

Buy article on-line for £20 (get immediate access)

Alerting Service

 RSS Feed

Permalink: http://dx.doi.org/10.1255/jnirs.850
QR Code (what is this?):

---