Noninvasive methods for dynamic mapping of microbial populations across the landscape

Soil microorganisms drive key ecosystem processes, and yet characterizing their distribution and activity in soil has been notoriously difficult. This is due, in part, to the heterogeneous nature of their response to changing environmental and nutrient conditions across time and space. These dynamics are challenging to constrain in both natural and experimental systems because of sampling difficulty and constraints. For example, soil microbial sampling at the Landscape Evolution Observatory (LEO) infrastructure in Biosphere 2 is limited in efforts to minimize soil disruption to the long term experiment that aims to characterize the interacting biological, hydrological, and geochemical processes driving soil evolution. In this and other systems, new methods are needed to monitor soil microbial communities and their genetic potential over time. In this study, we take advantage of the well-defined boundary conditions on hydrological flow at LEO to develop a new method to nondestructively characterize in situ microbial populations. In our approach, we sample microbes from the seepage flow at the base of each of three replicate LEO hillslopes and use hydrological models to `map back' in situ microbial populations. Over the course of a 3-month periodic rainfall experiment we collected samples from the LEO outflow for DNA and extraction and microbial community composition analysis. These data will be used to describe changes in microbial community composition over the course of the experiment. In addition, we will use hydrological flow models to identify the changing source region of discharge water over the course of periodic rainfall pulses, thereby mapping back microbial populations onto their geographic origin in the slope. These predictions of in situ microbial populations will be ground-truthed against those derived from destructive soil sampling at the beginning and end of the rainfall experiment. Our results will show the suitability of this method for long-term, non-destructive monitoring of the microbial communities that contribute to soil evolution in this large-scale model system. Furthermore, this method may be useful for other study systems with limitations to destructive sampling including other model infrastructures and natural landscapes.