The most common method of mapping grassland biodiversity comes from the spectral variation (SV) hypothesis. This method assumes that the variability in the spectral signal detected by the remote sensing instrument is correlated with biodiversity (Rocchini et al., 2004). This can be performed with both multi and hyperspectral instruments and can be used to assess biodiversity across a landscape and within plots from individual sites (Figure 2). Additional methods for monitoring both biodiversity and functional traits include the use of different spectral and vegetation indices, 3D data such as from LiDAR or structure from motion (SfM), data fusion between high spatial resolution imagery and that with greater spectral information, or spectral data combined with 3D (Aasan et al., 2015; Gašparović et al., 2019; Laliberte and Rango, 2011).
For this review, a Google Scholar search was conducted for studies between 2018 and 2024 with the search terms “remote sensing”, “grasslands” and “biodiversity”, and a separate search replacing “biodiversity” with “functional traits”. The publications that included all the search terms and were directly related to the topic were examined in further detail and their references checked for additional publications. This resulted in 37 publications, 20 related to biodiversity, 15 on functional traits, and two studies with a focus on both biodiversity and functional traits (Table 1).
Table 1: List of the main grassland biodiversity and functional trait remote sensing studies used for this review.