Using remotely sensed indicators of primary productivity to improve prioritization of conservation areas for top predators

Abstract

The need to incorporate habitat dynamics into spatial conservation prioritization is increasingly recognized but often limited by the lack of available temporal data. To overcome this limitation, model-assisted monitoring has been advocated and developed over the last decade. However, predictive modelling often relies on static habitat features or interpolated climate data, failing to consider other relevant ecological processes such as primary productivity and its dynamics. The importance of primary productivity relies on its key role in determining the availability of food resources via trophic cascades. This study aims to assess how the spatiotemporal variation in the habitat quality for top predators (using raptors as case study) can be affected by temporal fluctuations in primary productivity, and whether its incorporation into habitat modelling improves predictions in space and time. Through model predictions we then incorporate interannual habitat-quality dynamics into spatial conservation planning to identify priority areas for raptor conservation based on a new set of dynamic priority-area indices. Using raptor occurrence data from two surveys (2001 and 2014), we found high model performance and transferability for all species (AUC2001 = 0.89 ± 0.10 and Boyce’s Index2001 of 0.78 ± 0.20; AUC2014 = 0.89 ± 0.12 and Boyce’s Index2014 of 0.87 ± 0.06). Annual mean and minimum values of EVI (indicators of productivity and phenology, respectively) and elevation were the most important predictor variables for most raptor species. However, the standard deviation of EVI (an indicator of seasonality) was also found to be an important predictor for migratory species breeding in shrublands and croplands. Interannual model predictions showed a large interannual variation in habitat availability across species, partially associated with regional land-use changes. Interannual habitat-quality dynamics was translated into marked changes in the priority areas for raptor conservation —areas mostly associated with deciduous forest and water environments (rivers and dams) in our study area. Our results shed light on the role of intra- and interannual dynamics of primary productivity in driving habitat quality for top predators. They also confirm the potential of remotely sensed ecosystem functioning attributes to inform about habitat-quality dynamics in a cost-effective, standardized and repeatable way. Incorporating interannual dynamics of primary productivity in habitat modelling supports raptor conservation prioritization by informing about adequacy, stability and legacy of raptor conservation areas.