The radiometric indices of vegetation and soil can be used to synthesise annual variations by calculating Dynamic Habitat Indices (DHIs), which have shown good correlation with animal species richness at regional scales.
DHI uses a combination of vegetation indices, such as the Normalized Difference Vegetation Index (NDVI) to create a map that identifies areas with high habitat quality and connectivity. The map is then used to identify priority areas for biodiversity conservation. In this study, we hypothesized that the DHI used as indicator can improve the prediction accuracy in SDMs. We developed a comparative framework for two types of SDMs, based on predictors obtained from commonly used data (Land Cover classification - LC) and from remote sensing (RS) data, i.e. DHIs.
The study area used to demonstrate the approach, is the French region of Île-de-France, which is encroached by intensive agriculture and urban areas. We computed predictor sets based on classified LC and RS Sentinel-2 data. Initial analyses showed similar SDM prediction accuracy between the standard LC and RS based approaches. However, further analysis showed that the niches predicted by SDM from RS were larger than those predicted by SDM from LC. We determined that this difference was due to a threshold effect of binarization of potential presence scores on the distance variables to the LC classes.
DHI, as a predictor in SDMs, may therefore be more suitable than LULC predictors for identifying species niches on a regional scale, which can be used in biodiversity monitoring. Overall, DHI can be a valuable tool for conservation as an operational method, providing an objective and efficient way to prioritize conservation efforts and monitor changes in habitat quality over time.