Cross-scale monitoring of habitat suitability changes using satellite time series and ecological niche models

One of the biggest challenges to deal with the global crisis of biodiversity loss is the lack of efficient and viable monitoring systems across scales. Unlike traditional in situ biodiversity monitoring, a usually costly and time-consuming enterprise, satellite remote sensing (SRS) data offer a technically feasible and sustainable in time solution. Here, we devise a cost-effective and upgradeable spatiotemporal framework for monitoring the species-specific habitat availability changes across scales by trend analysis of habitat suitability index (HSI) de -rived from ecological niche models (ENMs; Maxent) and using time series of SRS data (MODIS). The SRS-ENM framework was applied for a large suite of native species (911), from major taxonomic groups (flora (vascular plants), amphibians, reptiles, birds and mammals), and listed in the IUCN Red List at regional (Iberian Peninsula) and continental (Europe) scales. The HSI-trend analyses predict cumulative reductions in habitat suitability for Threatened and Non-Threatened species across scales for the period 2002-2016. Specifically, 19% and 66% of the total grid cells for both species' groups showed negative trends at both regional and continental scales, re-spectively. Results were similar when considering all IUCN threat categories. All taxa groups showed a decrease in habitat suitability, but amphibians and reptiles groups hosted the largest number of negative HSI-trends grid cells. Considering all groups together, 12% and 34% of both study areas have strong reductions in habitat quality. We conclude that our framework detects increases and decreases in species' habitat suitability regardless of the spatial scale, extent, and pixel size. Species' range predictions across space and time based on SRS time series rep-resent a promising Earth observation tool to support traditional risk assessment protocols and anticipate the decision-making process, while serving as a cross-scale biodiversity monitoring system. (c) 2021 Elsevier B.V. All rights reserved.