Remotely detecting biocrust communities across an Antarctic terrestrial landscape

Background/Question/Methods

Autotrophic microbial communities are the primary drivers of carbon cycling in the McMurdo Dry Valleys (MDV) of Antarctica. Dense microbial mats, consisting mainly of cyanobacteria and moss, occupy aquatic areas, while microbial communities also occur at much lower densities as surface soil biocrusts across the terrestrial landscape. Our previous research has shown that multispectral satellite data can be used to detect and quantitatively estimate microbial mat biomass in high-density areas like the riparian zones of streams and along lake margins, where Normalized Difference Vegetation Index (NDVI) approaches have been quite successful at quantifying the abundance of surface microbial communities. Given the broader spatial extent of the arid terrestrial landscape outside of stream channels and lake margins, we predict that a significant proportion of the valley-wide ecosystem carbon budget is represented by these patchily distributed biocrust communities. Here, we describe recent remote sensing and field-based survey and sampling approaches to describe the distribution of biocrust communities in “dry” soil environments. This work is part of ongoing efforts to refine the carbon budget for this region and to examine controls over the distribution and activity of these critical soil communities.

Results/Conclusions

We have found that spectral parameters like NDVI applied to WorldView-2 (WV-2) satellite imagery can be influenced by lithological surface composition in addition to photosynthetic biology. Therefore, we implemented additional multispectral and hyperspectral tools to quantify the detectability of biocrust in this region. Using the eight WV-2 multispectral reflectance bands and traditional vegetation indices (e.g., NDVI, SR, SRre, and NPCI) in a principal components analysis, we documented distinct patterns between known biotic and abiotic surfaces from study plots in the field. Currently, this approach has a detection threshold of ~ 20 mg/cm2 ash-free dry mass of active biocrust and extends the capability of remote-sensing of microbial communities out of the aquatic margins and into the dry soils. We also applied linear spectral unmixing models to WV-2 data using lab-derived spectra of biotic and abiotic materials, and additionally determined that in a lab setting using hyperspectral data, we can confidently detect biocrust abundances down to 1% (g/g). This interdisciplinary work is critical for measuring and monitoring terrestrial carbon stocks and predicting future ecosystem dynamics in Antarctic ecosystems which are particularly climate-sensitive and difficult to access.