AI applied to an early warning and detection system for Phytophthora cinnamomi in cork oak forests/shrublands

BACKGROUND

Strengthening scientific and technological cooperation in priority areas for border areas is important for both Portugal and Spain, with a view to enhancing and sustainably developing the environmental, economic and societal resources of these regions.

In 2020, within the framework of the Mission, Vision and Values it pursues, the la Caixa“ Foundation” launched, for the third year, the Promove Program - The Future of the Interior, in partnership with the BPI and Foundation for Science and Technology. The aim is to support innovative initiatives in strategic areas for the development of Portugal's inland regions, which can be replicated in other regions with similar characteristics.

The application of monitoring and warning systems in large forest patches, using digital robotic and artificial intelligence (AI) technologies on a large scale, was the theme within which InnovPlantProtect (INPP) framed this two-year project, which was one of the winners of the 2020 edition of the Promove Program: “AI applied to a prediction and early detection system for forest fires". Phytophthora cinnamomi in dehesa ecosystems/ dehesa”.

SYNOPSIS

The Portuguese montado is an agropastoral system similar to the Spanish one dehesa, both recognized for their cultural, economic and ecological value. The cork oak (Quercus suber) and the holm oak (Quercus rotundifolia), two of the representative species of this cross-border ecosystem (Alentejo, Extremadura and Andalusia), allow us to enjoy various services, such as the production of cork and firewood, food for animal species (black pigs and cattle), carbon sequestration, reduction of soil erosion, habitat for rare and endemic species, as well as being important recreational spaces.

In Portugal and Spain, the montado/ dehesa has been suffering a worrying decline for several years. The cause of this decline is not always clear and is due to various factors, such as high water stress, poor nutrients in the soil, overgrazing, poor land management and the presence of pests and diseases. One in particular, the oomycete Phytophtora cinnamomi (Pc), a parasitic fungus-like organism introduced to Europe at the beginning of the 18th century, has been identified as one of the main pathogens.

Recent studies based on remote sensing techniques, spatio-temporal epidemiological analysis at the landscape level and assessment of fragmentation and connectivity in forest ecosystems have contributed to a better understanding of the action of the Phytophtora cinnamomi as the cause of the disease known as cork oak decline.

OBJECTIVES

The proposed project aims to use the latest digital remote sensing and Artificial Intelligence (AI) technology to develop an early warning and detection system and spatial transmission models for PC outbreaks.

There are three study areas: two in Portugal (Avis and Ourique municipalities) and one in Spain (Alcuescar). Two producer associations are involved in the project as stakeholders: a ACPA - Association of Alentejo Pig Breeders and the Aflosor - Agro-Forestry Association of the Ponte de Sor Region.

ACTIVITIES

In order to achieve the planned objectives, the project takes the form of various activities. Some are carried out on the ground:

1. inventory of trees and visual assessment of disease symptoms;

2. soil and root sampling;

3. calculating the leaf area index;

4. soil sampling for analysis of physical and chemical characteristics;

5. drone campaign;

6. participation in events.

Other activities are carried out in the experimental laboratory (identification of the pathogenic organism) and in the computer lab (image analysis and development of AI-based models).

ORGANIZATIONS INVOLVED