Using earth observation data to improve climate-modifying habitat datasets for pest risk modelling

Slides presented at the International Pest Risk Research Group (IPRRG) Annual meeting in Nairobi, Kenya 2023:

Using earth observation data to improve climate-modifying habitat datasets for pest risk modelling. Tim Beale (CABI), Pascale Bodevin (CABI), Steve Edgington (CABI), Libertad Sanchez Presa (CABI), Bryony Taylor (CABI), Alex Cornelius (Assimila), Gerardo Lopez Saldana (Assimila), Jon Styles (Assimila), Darren J. Kriticos (Cervantes Agritech)

Non-climatic habitat factors can have a significant effect on species ranges, allowing them to persist well beyond their natural ranges.  Irrigation and protected agricultural structures such as glasshouses are used specifically to allow crop species to be grown successfully in locations where the climate is otherwise inhospitable.  The same conditions that allow the crops to be grown in hostile climates allows pest species to persist beyond their natural limits.  Species distribution databases such as GBIF and iSCAN do not distinguish between species distribution records collected from natural habitat situations and those from artificial habitat.  Bioclimatic models that ignore the role of these artificial habitat modifications routinely make egregious errors, incorrectly projecting habitat suitability into inclement climates.  Methodically overestimating the pest risk area in this manner can have important effects on biosecurity risk management, misdirecting resource allocation for preparedness activities, and undermining the reputation of pest risk assessment.

Advances in Earth Observation (EO) technology have opened up new possibilities for addressing agricultural challenges in the face of climate change. The EO4AgroClimate project is using EO-derived data to enhance three critical modelling datasets: irrigation, protected agriculture, and canopy temperature. These datasets will help to contextualise species distribution data from repositories, as well as improve the performance of environmental niche models (ENM) leading to more accurate, high-resolution, and timely information for pest risk assessment.