Deciphering kiwifruit seed germination using neural network tools

J.E. González-Puelles, M. Landín, P.P. Gallego, M.E. Barreal
We are interested in obtaining rootstocks from seedlings of different Actinidia species such as Actinidia chinensis, Actinidia arguta and Actinidia eriantha, but the literature about kiwifruit seed germination is scarce and generally reports poor germination rates. The aim of this work was to establish what factors are critical to kiwifruit seed germination. Germination is a very complex process, controlled by multiple factors including temperature, dormancy, water availability, and light. The use of cold temperatures, or plant growth regulators (gibberellins) to substitute the chilling requirements, to break kiwifruit seed dormancy has been already described. Here, several factors such as cold stratification (time and type of treatment) and thermo-photoperiod conditions were tested. Response parameters percentage of seed germination (%G), mean germination time (MGT), and germination index (GI) were calculated. In this study, we present results from A. chinensis var. deliciosa 'Summer' in order to show how neural networks can be useful to decipher key factors of the germination process. Neurofuzzy logic models simplify data analysis and point out the critical role of cold stratification time (long periods at 4°C) and stratification treatment (using gibberellic acid) on kiwifruit seed germination.
González-Puelles, J.E., Landín, M., Gallego, P.P. and Barreal, M.E. (2018). Deciphering kiwifruit seed germination using neural network tools. Acta Hortic. 1218, 359-366
DOI: 10.17660/ActaHortic.2018.1218.50
https://doi.org/10.17660/ActaHortic.2018.1218.50
Actinidia chinensis var. deliciosa, thermoperiod, photoperiod, seed dormancy, seed stratification, plant growth regulators
English

Acta Horticulturae