Application of artificial neural networks for ecological niches modelling in Araucaria angustifolia and Ilex paraguariensis

• Grant number: 18/22777-0
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: February 01, 2019
• End date: January 31, 2020
• Field of knowledge: Biological Sciences - Genetics - Plant Genetics
• Principal Investigator: Karina Martins
• Grantee: Gustavo Henrique Sousa Rodrigues
• Host Institution: Centro de Ciências Humanas e Biológicas (CCHB). Universidade Federal de São Carlos (UFSCAR). Sorocaba , SP, Brazil

Abstract

Nowadays climatic changes represent the main focus of discussion on the conservation of natural populations. As trees have a long life cycle, adaptive selection processes occur more slowly than the speed of climatic change. As a consequence, typically cold-weather species such as Araucaria angustifolia and Ilex paraguariensis may suffer a range reduction and, in some more drastic climate scenarios, even the extinction of entire populations. In this context, developing breeding and genetic conservation programs toward adaptation to future projected climates is an interesting strategy. Ecological Niche Modelling (ENM) methods are helpful to identify geographic regions where species grow better, thus providing valuable information for breeding programs. Artificial Neural Networks is an ENM method that results in niche models with high accuracy and efficiency. Thus, the objective of this project is to use Artificial Neural Networks to model regions where A. angustifolia and I. paraguariensis grow better, considering present and future climate scenarios. Two phenotypic variables (total height and diameter at breast height) and from six natural locations of occurrence and several climates and edaphic data will be used for the ANNs. These will be constructed using the technique Multilayer Perceptron, in software R, where the number of neurons and layers will be tested. The best models will be used to map the regions corresponding to ecological niches of A. angustifolia and I. paraguariensis. These results will provide an indication of populations that have the potential to be used in conservation and breeding programs of A. angustifolia and I. paraguariensis in a scenario of climate change. (AU)