Measuring plant diversity using spectral imaging

We have known for decades that the chemical composition of plants can be estimated from reflectance spectra. What we found is that the spectral dissimilarity, or the overall differences in spectral reflectance, among plant species increases with their functional dissimilarity and evolutionary divergence time.

The value of ecological biodiversity for maintaining ecosystem stability and function is well established, but how do we measure it at larger scales. We need novel approaches that are rapid, repeatable and scalable in particular in ecosystems for which information about species identity and the number of species is difficult to acquire.

A group of US researchers is proposing to measure plant diversity using spectral data in an attempt to  improve efforts to predict how well ecosystems function. The colleagues measured the light reflectance of plants in 35 plots at a field station north of Minneapolis famous for long-term ecological experiments by using a field spectrometer. The spectrometer allows the researchers to evaluate how much light plants reflect at the leaf level across a range of wavelengths. By taking the leaf-level data the team found that the spectral diversity of a plant community predicted aboveground productivity, a critical ecosystem function, to a similar or higher degree than measures of species functional differences, their phylogenetic distances or species richness in a plant community.

Seeing that the ecosystem effect of plant diversity can be effectively evaluated using spectrometry, the team also wanted to know if their method could scale. They used an imaging spectrometer mounted three meters above ground at the same 35 plots at Cedar Creek. Their scans showed that the spectral diversity metric performed similarly when calculated from such spectral images.

The findings indicate that spectral diversity provides a powerful, integrative method of assessing several dimensions of biodiversity relevant to ecosystem function. The rapid changes in the Earth's biodiversity that are underway require novel means of continuous and global detection. This study demonstrates that we can detect plant biodiversity using spectral measurements from plant leaves or from the sky, which opens a whole new range of possibilities.

I guess all that needs to be shown is how well it really scales when it comes to remote sensing technology but this is really promising especially when taking into account the breadth of additional information the colleagues were able to obtain.