Multi-species near infrared calibration models for predicting nitrogen status of ornamental cuttings – strategies for scope extension (Dieter Lohr)
Nitrogen status is a well-known key factor for adventitious root formation of ornamental cuttings. Even a slight nitrogen deficiency of stock plants impairs subsequent rooting of cuttings, whereas a surplus can either decrease or increase it. The effect of nitrogen surplus depends on carbohydrate reserves at harvest date and current photosynthetic activity of cuttings during rooting. However, nitrogen status of cuttings cannot be used as parameter for quality assessment of cuttings yet, because established wet-chemical methods are too time-consuming and labor intensive. Near-infrared spectroscopy (NIRS) might be a reliable alternative, especially if any sample preparation, such as drying and grinding, is avoided. The major obstacle for practical use of NIRS is that quite high number of samples is needed to develop robust calibration models. Depending on the complexity of the matrix and the analytes of interest, several hundred samples have to be measured by NIRS and a reference procedure. Thus, the aim of this research is to evaluate strategies to extend existing calibration models to new ornamental species with a reduced number of samples.
Firstly, joint calibration models for four N pools (soluble, insoluble, total organic and total N) were built with about 200 pelargonium and chrysanthemum cuttings each. Subsequent, strategies to extend these models to impatiens, osteospermum, poinsettia and petunia with a small number of samples were tested. Applying the calibration models directly to the four new species results in increased systematic errors (BIAS) whereas random errors (SEPC) were similar to those of pelargonium and chrysanthemum. Stepwise addition of up to 40 samples per species to the calibration set significantly reduced BIAS. On the other hand, reducing the number of pelargonium and chrysanthemum samples had no clear effect on BIAS and SEPC.
Overall, NIRS has been proven to be a valuable tool for analyzing nitrogen status of ornamental cuttings, which can be used for quality assurance. Besides the findings for nitrogen pools present in the current research, encouraging results were also obtained for leaf carbohydrates (especially starch and total non-structural carbohydrates). Future research will focus on the implementation of the technique in production processes of commercial stock plant facilities.
Dieter Lohr won an ISHS student award for the best poster at the III International Symposium on Horticulture in Europe (SHE2016) in Greece in October 2016.
Dieter Lohr, Weihenstephan-Triesdorf University of Applied Sciences, Institute of Horticulture, Am Staudengarten 14, 85354 Freising, Germany, e-mail: email@example.com