Modeling of plant tissue culture components using artificial neural networksa case of actinidia arguta

Abstract

The kiwifruit is a woody plant native to China, belonging to the Actinidiaceae family, genus Actinidia, whose fruit is cultivated in almost all continents, and whose production has been steadily increasing in the last 10 years. It comprises 54 species and 21 varieties, with Actinidia chinensis var. deliciosa cv. "Hayward" being the most cultivated, exported and sold. The species used in this doctoral thesis is Actinidia arguta (Sieb. et Zucc.) Planch. ex Miq., also called "hardy kiwi", "baby kiwi" or kiwiño, characterised by a fruit with a trichome-free, smooth and edible skin. The size of the fruit depends on the cultivar, although they are usually small. This kiwifruit species originates from East Asia. A special feature of the kiwifruit cultivar used in this thesis (Issai) is its monoecious character, which allows self-fertility. In order to obtain a large amount of plant material of this species in a relatively short period of time, the use of micropropagation techniques was chosen, which allow a rapid clonal propagation of genotypes of interest from explants. In a first part, the effect of plant growth regulators, culture medium and subculture on shoot number and shoot length will be evaluated using artificial neural networks (ANN). In a second experiment, the determination of key mineral nutrients and the prediction of the best combination of salts for the correct micropropagation of kiwifruit will be considered using computer tools such as design of experiments (DoE) and ANNs. Once this part has been completed, the thesis will focus on deciphering, by means of neurofuzzy logic, the effect of mineral nutrients and, above all, vitamins, since it is known that these organic compounds have various physiological functions, although their role in the development of micropropagated plants is unclear and controversial. Finally, the effect of mineral nutrients, vitamins and plant growth regulators on kiwifruit micropropagation will be assessed with the aim of deciphering and generating new knowledge on the hierarchy of the components of the culture medium for the subsequent formulation of new basal media, which will improve the quality of micropropagated woody plants.