Machine learning tools to identify key minerals of Hoagland solution for healthy kiwiberry micropropagated plant hardening

S. Maleki, B. Maleki-Zanjani, B.B. Kohnehrouz, M. Landin, P.P. Gallego
Hoagland’s solution is widely used for growing plants in hydroponic systems. The objective of this work was to know in depth the effect of mineral elements in Hoagland’s solution on various morphophysiological responses of growth and quality of micropropagated kiwiberry seedlings during their acclimatization stage. For this, a five-dimensional experimental design was established based on 19 model points with optimal response, which included the combination of 7 mineral elements (N, P, Ca, K, S, Mg and Cu) at three levels. Subsequently, a model was built using a machine learning tool, called neurofuzzy logic, which made it possible to understand the role of each mineral element and determine how the mineral components of the tested solution affected physiological responses. The results show that the combination of both computer tools was able to predict and unmask the hidden interactions between minerals. Halving the nitrate content of the full concentration Hoagland solution is essential to improve the acclimatization of micropropagated kiwiberry plants.
Maleki, S., Maleki-Zanjani, B., Kohnehrouz, B.B., Landin, M. and Gallego, P.P. (2022). Machine learning tools to identify key minerals of Hoagland solution for healthy kiwiberry micropropagated plant hardening. Acta Hortic. 1332, 39-46
DOI: 10.17660/ActaHortic.2022.1332.6
https://doi.org/10.17660/ActaHortic.2022.1332.6
Actinidia arguta, artificial intelligence algorithms, ex vitro acclimatization, design of experiments, physiological disorders
English

Acta Horticulturae