Postcard from a Fellow: Daniel Smith’s summer obsession with flumes, fake roots, and Psych

By Daniel Smith |  August 9, 2020

Hi GCC community and friends! Daniel Smith here; I’m a fourth year PhD candidate within the Biological Systems Engineering (BSE) Department and I’ve been an IGC fellow since spring 2019. While at Virginia Tech, my research has focused on how plant roots protect streambank soils from fluvial (water) erosion.  I am interested in understanding which plant root processes/mechanisms have the most influence on streambank erosion. I’m excited to share some of the tools I have been using and testing out this summer that will be used to measure soil erosion in my future experiments.

First, let me introduce the epic, room-sized (26 ft long by 3 feet wide) flume. This flume, housed in ICTAS II (the Institute for Critical Technology and Applied Science), was designed and built to represent water flow within a stream channel. However, a major distinction must be acknowledged between the manmade flume and a natural streambank. The bottom and sides of this flume are made out of smooth, plexiglass material while streams typically have rough bed sediments and grainy bank soil. Many streams also have visible plant roots growing along the streambank face, adding an extra layer to the grainy soil material.  Consequently, to measure the effect of plant roots on streambank erosion in the flume, I need to better represent the boundary conditions found in natural streambank settings.

The flume I am using for my future experiments. Currently, two of the four flume inserts are sitting inside the flume. Photo credit: Trerese Roberts

Part of this modification was already done from a previous student’s research project. That student built four “flume inserts” made out of a wooden frame and 1-inch thick PVC sheeting. Sand particles were glued onto the PVC sheets, giving them a grainy texture, and a hole was cut into one of the panels so a soil sample could be placed there for erosion testing. For my experiment, I worked with Allen Yoder in the BSE department to make the testing holes larger, replace some broken and/or missing parts, and fix any worn out sections of the frame. Once inside the flume, these updated walls would represent streambank soil that had no vegetation.

Soil aggregates attached to stream bank plant root surfaces
Up close look at two flume inserts that needed further modifications before being put back into the flume. Photo credit: Trerese Roberts

While the sand wall is a good representation of an unvegetated streambank, I still needed something that matched vegetated streambank soils with roots facing the stream channel.  As a result, more PVC sheeting had to be purchased, cut to the correct size, and covered in sand particles. To represent the roots of grassy plants, I decided to use different diameters of flexible, 100% polyester thread. Here’s the fun part: given the size of my PVC sheets, field data from another study revealed I would need ~1500 roots glued onto each insert in order to match what was typically found in the field!  Armed with a hand drill, scissors, thread, and E6000 glue, I drilled 200 holes into each PVC sheet, cut and tied different thread diameters together, and individually glued these fake root bundles into each hole. Once complete, these walls can be drilled on top of the sand wall inserts when I am testing rooted soil planted with grassy-type vegetation. Between the tediously long hours of cutting and gluing, I’ve been able to watch multiple documentaries and an unknown amount of the show Psych!

So far, only one of the four new rooted PVC sheets is nearly complete. It takes about two or three full days to finish one. Photo credit: Trerese Roberts.

But wait…there’s still more! The sand walls will be used to test erosion in unvegetated soil samples and the polyester threaded walls will be used to measure erosion in soil samples planted with grassy vegetation. What about the woody plants?  You guessed it– I’ll need to make another set of walls with fake root material that represents woody (e.g. more rigid) roots. Once that task is completed, I will run some preliminary tests in the flume to make sure the walls are working as desired before the real experiment starts. Needless to say, by the end of this summer, it’s likely that I’ll have watched so many episodes of Psych some of them will start blending together in my head…

Closeup of "fake roots" of PVC sheets. Photo credit: Trerese Roberts
Switchgrass and dogwood plants growing nicely in the greenhouse. Photo credit: Trerese Roberts

In addition to the glorious and time-consuming task of gluing near 800 root bundles (~6000 fake roots in total), I am also taking care of some real plants this summer in the Hahn Horticulture Garden greenhouses. My future experiment will look at using switchgrass (Panicum virgatum) and silky dogwood (Cornus amomum) to represent grassy and woody plant roots. This summer I am testing to see how they grow in these PVC pipe chambers to see if any modifications will need to be made later. So far things are growing nicely!

Daniel Smith is an Interfaces of Global Change fellow working with Dr. Tess Thompson in Virginia Tech’s Department of Biological Systems Engineering. He is studying how plant roots and soil microorganisms impact streambank soil resistance to fluvial erosion.