I recall hearing a business maxim some years ago: “What gets measured gets done.” In other words, if there is no standard by which to measure something, it becomes a low priority. Unfortunately, this seems to be the case with soil erosion.
For years, resource professionals have had the RUSLE2 soil erosion model to predict the amount of sheet and rill erosion that occurs on a farm. Sheet and rill erosion is caused by a shallow sheet of rainwater that runs across the land. This picks up and transports soil particles already detached by a raindrop splash or by soil disturbances such tillage operations. Because there was a way to estimate sheet and rill erosion, soil scientists set standards, or benchmarks, for it. Using the RUSLE2 model, soil scientists developed the standard for “T” or the “Tolerable soil loss” level. This “T” is the erosion level that Natural Resources Conservation Service (NRCS) encourages farmers to achieve. The tolerable soil loss for most soils in the Midwest is set at 5 tons/acre/year. As with all standards, experts argue whether it should be higher or lower but the fact remains, resource professionals and farmers have a method to estimate how changes to farming practices impact soil erosion. In other words, we can set standards and measure progress.
Unfortunately, the same cannot be said for ephemeral gullies. Ephemeral gullies are defined as small channels eroded by concentrated flow that can be easily filled by normal tillage, only to reform again in the same location by additional runoff. Ephemeral gully erosion is far more visible than sheet and rill erosion and everybody can see it is occurring. But still, resource professionals have no method to predict the amount of ephemeral gully erosion that occurs at a field level. Ephemeral gully erosion is believed to be at least as significant as sheet and rill erosion in terms of sediment delivered from cropland to streams, rivers, and lakes. In fact, a special report written jointly by the USDA Agriculture Research Service (ARS) and NRCS emphasized the need to focus on ephemeral gully erosion. This report estimates that 40 percent of watershed sediment yield can be attributed to ephemeral gully sources (Bernard, 2009).
To date, there is no model similar to RUSLE2 that can be used to estimate or predict the level of ephemeral erosion occurring in a farmer’s field. Therefore, ephemeral erosion has always been the ugly step sister to sheet and rill erosion. With no way to predict ephemeral gully erosion, it has received far less attention, and no standard for ephemeral gully erosion reduction has ever been set.
After all of these years, finally, USDA-ARS is developing a model that can be used to measure ephemeral erosion. Dr. Seth Dabney is working to develop a model that can be used to predict the location and the amount of ephemeral erosion. Hopefully, resource professionals will take notice of Dr. Dabney’s work and get serious about measuring and reducing ephemeral erosion.
Not having an ephemeral gully erosion calculator is a disservice to all of those working to protect natural resources. Remember, precision conservation is about applying conservation practices in the right place, at the right time, and at the right scale so we can solve the right problem. How can resource professionals be expected to apply the right practice if they don’t know how much ephemeral gully erosion is occurring? In some cases, we have been applying practices to reduce sheet and rill erosion when we should have been looking at ephemeral gully erosion.
Great points, Tom. Strong evidence exists that tools used today to make soil erosion estimates give answers much below what is actually occurring, as you point out. Further, increasing evidence exists that what we have been considering tolerable soil loss is much greater than soil renewal rates, and what is actually sustainable. Designing farming systems to meet a soil loss rate, T, that is admittedly not sustainable, and then using a system of tools that consider only part of the soil erosion occurring so that we can on paper meet T, is leading us down a soil resource depletion schedule rather than a soil sustainability plan.
A great article…one that I fully endorse and I wish we could get around tolerable. I think we should move to the S soil lost rate…Sustainable. There is so much more erosion out there than just sheet and rill erosion. If we don’t pay attention to the other 50% as you call it (perhaps more?), soon we won’t have the topsoil to raise our crops…then what? I am all for production and making money, but at what cost?
People (not just farmers) need to understand that erosion is pollution. Erosion causes more problems than dirt in the ditch or a gully that we have to fill in each fall. It affects our future and our kids future. If we don’t pay attention to that, then what? There is ephemeral erosion, shoreline erosion, streambank erosion, construction site erosion, and I could go on. In this state, the biggest erosion problem we have is, regretably, agricultural in nature.
If there is anything I can do to help you promote this, let me know.
I absolutely agree we need to do more work to properly calibrate RUSLE2. Like RUSLE2 we need to create new tools to assist us in identifying the source and solutions for other types of soil erosion. Precision conservation requires that we understand where the erosion is occurring and understand the most cost effective practices that can be implemented to reduce that erosion.
If the goal is to reduce the silt entering a lake or stream, it is crucial to know where to focus our attention. If the silt is coming from sheet & rill erosion we can use RUSLE2 to have identify potential solutions. If the erosion is coming from ephemeral erosion we currently do not have any tools to help producers determine the amount of silt leaving their farms and the cost/benefit of installing practices like grassed waterways.
I hope your work along with Dr. Seth Dabney (on your USDA NIFA grant) will help shed some light on new methods to estimate ephemeral erosion.
I always appreciate your insights on matters such as this. Your rare perspective that combines business, science and conservation is needed in the conservation arena. I remember that Leonard Johnson, a soil scientist here at Madison who retired 20 years ago wrote an article in the JSWC where he argued that legitimate T values should be 1/10 of what we are using. It is clear we don’t have enough resources to focus on all instances of excessive soil erosion, but as you point out, missing ephemeral gully erosion really challenges us to target our limited resources on the critical areas. Yes, lets develop the tool to measure what is really happening on our agricultural lands, and then when the enormity of that challenge confronts us, lets hope there is the wisdom and political fortitude to target our limited resources.
Thanks for raising this issue. It turns out ephemeral gully erosion is harder to predict than the sheet and rill erosion that RUSLE2 and WEPP predict.
From the modeling perspective, there isn’t a large enough database to support the kind of statistical approach that underlies RUSLE2 central tendencies. Therefore, we are trying to develop a process-based representation. Tillage moves soil to fill and reset gully dimensions while distributing the degradation to areas beyond the ephemeral channel. Tillage erosion needs to be accounted for as well if we want to talk about sustainable levels of soil erosion.
NRCS has not made ephemeral gully modeling a high priority. This may be because a conservationist could go to a field and measure the dimensions of a gully and thus know how much erosion had occurred. Therefore, they didn’t need a model to tell them how big their problem was the same way that they did for sheet and rill erosion. If they saw gully erosion, they knew they had to do something about it and could do a cost/benefit analysis based on the size of the gully in the field being considered. However, when we become concerned of off-site impacts, then being able to predict gullies over larger areas takes on new importance. Unfortunately, many of the tools currently being used to do this are really inadequate. Reliable ephemeral channel erosion estimates cannot be simply based on the use of a different rainfall/runoff erosivity term or by using a channel C factor.
The channel erosion component used in the WEPP watershed version is a process-based model that predicts first downcutting to a non-erodibile layer and then widening. That was our starting point for the model we are developing. The WEPP channel model needed a great deal of refinement to be applicable to the complex network of gullies identified in the high-resolution digital elevation data available from LiDAR. The fact that the ephemeral gully channels have sizes that change with time makes this a challenging undertaking! Erosion/deposition is calculated by the increase/decrease of channel width and/or depth through time. Frequently, erosion may occur by the movement a series of knick points or the sloughing of side walls, but these details are usually simplified by assuming a uniform friction slope and calibrating erodibility and critical conditions as based on observed sediment delivery from simulated controlled channel experiments. To calculate ephemeral gully erosion, we need adequate representations of runoff rates, soil erodibility (that varies with soil depth and management), the size and density characteristics of eroded/transported sediment, and sediment transport capacity relationships (based on shear stress or stream power) that can be applied to the different sediment fractions. Flow and sediment transport/deposition must be apportioned with the eroding channel and the surrounding floodplain (field area adjacent to the gully channel). The tool we need must route the runoff and the sediment downslope and must reflect the influence of backwaters caused by changes in roughness or grade control structures. Our model development progress has been slower than desired, but this is a big job and we are systematically making progress.
Our ongoing Iowa project will allow us to test our numerical solution for a few specific fields. This will be a very valuable start. To solve the problem nationally, we need even more testing/validation since our tool will be used for conservation planning without further calibration.
Seth & Rick,
I am excited to see your early measurements of ephemeral erosion on the Iowa sites (through the NIFA grant with Iowa State). It should help us get a better handle on the levels of ephemeral erosion that is occurring.
I think we could jump start the process of measuring gully erosion if we look outside of just one agency and their needs. While attending the Conservation Tillage Conference at the Ohio Northern University in Ada, Ohio, I was able to hear speakers who talked about tillage erosion work that the Agricultural Research Service (ARS) had done in the Western U.S. The speaker was able to put us in contact with researchers who could help us in our erosion studies on the end moraines in eastern Indiana. They also provided us access to research papers that had been written on the areas of tillage erosion.
I was happy you mentioned tillage erosion in your comment. Within the last week Dr. Seth Dabney (USDA-ARS) told me about tillage erosion. He said there is significant research to support tillage erosion and it is a significant part of erosion; maybe as significant as sheet and rill erosion. What I found most interesting from Dr. Dabney’s discussion is that RUSLE2 is capable of modeling tillage erosion. Unfortunately this feature in RUSLE2 goes largely unused in erosion predictions. I assume that tillage erosion is does not have a lot of effect on water quality. However I would think it would have a significant effect on soil health and therefore crop management. This is something I want to write about in one of my next posts. Thanks for bringing up tillage erosion. It is a very interesting topic for soil health.
I will follow-up with you on the research papers you mentioned in your comment.