While at the University of Leeds, I had the opportunity to investigate a naturally formed peat pipe, effectively an underground stream, in a blanket peat bog in the Pennines of Northern England. These natural pipes convey around 15% of the water in these catchments but are responsible for a disproportionally high amount of the particulates and carbon fluxes out of these landscapes, due to lots of erosion happening underground.
I was curious how water moved from the surrounding peat into these natural pipes, as the pipeflow varied a lot through time despite the surrounding peat at that depth supposedly being constantly waterlogged and nearly impermeable. Was the pipe supplied by a network of smaller pipes and macropores in a 3D version of the dendritic pattern seen supplying the surface stream?
To test this hypothesis, we excavated a pipe and carefully collected small cubes of adjacent peat. Once back in the laboratory, we measured how quickly water could pass through these cubes in different directions, learning more about the structure of macropore connections inside the peat.
We found the pipe itself was full of roots and that the hydraulic conductivity varied enormously between different samples of peat, and through the same sample when measured in different directions. The rates water could move through the peat varied by up to 7 orders of magnitude (~10,000,000 times), showing how difficult it is to come up with a meaningful average value of hydraulic conductivity when peat pipes can be so important. We were unable to tell whether the pipe had formed underground, or may have been a surface stream buried by the formation of new peat over the top. There are many more questions here for others to answer!
We published a report on this study in the Journal of Water Resources Research (Cunliffe, A., A. Baird and J. Holden, 2013, Hydrological hotspots in blanket peatlands: Spatial variation in peat permeability around a natural soil pipe, Water Resour. Res., 49(9):5342-5354. DOI:10.1002/wrcr.20435. Download PDF)