Biological soil crusts and global change: Spectrally monitoring moss responses to future climate change scenarios
Dryland biological soil crusts - a community of mosses, lichens, cyanobacteria, and heterotrophs living at the soil surface - are a fundamental component of the structure and function of arid and semiarid ecosystems. These soil communities play critical roles in dryand carbon fixation, nitrogen fixation, and soil stabilization, and existing data suggest biocrusts are sensitive to seemingly subtle changes in climate. For example, previous research on the Colorado Plateau showed dramatic mortality of the common moss Syntrichea caninervis in response to an increase in the frequency of small monsoonal rainfall events. Yet, despite the importance of these biocrust organisms, our ability to monitor biocrust responses to altered climate remains limited. Spectral imaging represents an under-exploited tool for documenting change within biological soil crust communities. Here, we induced stress within Syntrichia caninervis samples by increasing the frequency (twice weekly) of small (1.2mm) rainfall events, and used spectral analyses to monitor the moss’s progression towards eventual mortality. We focused on a number of indices as potential tools for assessing change, including the Normalized Difference Vegetative Index (NDVI). In addition, we concurrently examined shifts in nitrogen cycling within the soil matrix of stressed mosses to link moss stress, spectral imaging, and biogeochemical consequences.
As expected, mosses were strongly, negatively affected by the increased frequency of small rainfall events, and exhibited clear signs of chlorisis - a yellowing and reddening of moss leaves. Belowground changes to biogeochemical cycling occurred in the form of decreasing NH4+ concentrations, and concurrently increasing NO3- concentrations, alluding to a steady progression towards NO3- dominance in the soil. These changes to soil N cycling occurred before significant moss stress became visibly apparent, implying that even early stress to moss has large implications for soil fertility. We found spectral analyses an effective tool for quantify the progression towards chlorosis within this moss species, however, the NDVI wide band index that is widely used for remote sensing of vascular plants was an unexpectedly poor indicator of stress. Instead, narrow band hyperspectral indices were much more effective at quantifying chlorosis. These findings suggest that the correctly employed hyperspectral images could be a valuable tool for documenting change within biological soil crust communities. Tools such as these, which can document stress, will become increasingly important to dryland ecosystem assessment in light of the dramatic biogeochemical changes associated with moss mortality.