Wind Erosion Prediction System (WEPS)
Short Term Fix for High Organic Soil-Summary
February, 11, 2009
Background
The current version 1.0 of the WEPS does not run for high organic soils. NRCS must have a wind erosion prediction model capable of predicting the wind erodibility of high organic soils. NRCS soil records for high organic soils lack data needed to run the model. NRCS soil scientists took the position that the mineral fractions were so low in mucks that the land management decisions should not be based on them. ARS took the position that the mineral fraction and sand content in particular, is a critical driver of the wind erosion process. Thus the mineral fraction of the organic soils is not populated and WEPS will not be able to predict wind erosion on high organic soils because the mineral properties are not populated in the soil database. Certainly there is a large bulk density difference between an organic soil and a mineral soil.
It was decided that the ARS and NRCS should review high organic soils from the field perspective. Two trips were made to gather samples and review the field wind erodibility of high organic soil in Michigan and Florida.
The organic soils in Michigan were very erosive and soil loss was extensive. Local farmers and NRCS report that it is common for ditches to fill with wind eroded material. The row crops grown on the Michigan organic soils produce very little surface residue, soil surface structure, or soil crusts. All but one site had fairly high amounts of mineral soil (sand). NRCS found that the soils have changed substantially since they were mapped. The organic soils are much shallower now than when mapped. Some are no longer mucks because of the decrease in organic matter and the shallow depth of the organic layer. ARS realized that the organic portions of the field could move at a much lower threshold wind velocity than the mineral soil. Work is needed from both NRCS and ARS to characterize the high organic soils and develop the technology within WEPS to predict wind erosion.
The Florida soils were much less erodible and had only minor wind erosion issues, partly because the cropping systems provide adequate vegetative cover during the erosion season. The current Florida high organic soils are now much shallower than mapped compared to the Michigan high organic soils, due to oxidation and erosion. The Florida soil surface has much more structure and tended to crust when dry.
Both states have issues with subsidence. Most high organic soils, farmed in a fully drain condition will loose about one inch of depth per year due to oxidation-reduction. Florida regulates the water table to slow phosphorus movement into surface water. This has slowed the subsidence to about ½ inch per year.
Purpose
This paper will discuss a solution to allow the current WEPS model to estimate soil loss from all organic soils until long term research can be done and the new findings added to the model. The purpose of this paper is to review the available soil data and select a standard soil record to use in WEPS when making a soil loss estimate on organic soils.
Original Work
Figure 1 below gives us a basis to select the best fit soil file for organic soils. It would be better to consider using the average mineral soil file and a base management to set a baseline. The flat, wide, no crop comparison suggested at the FL meeting caused the model to predict over 100 ton/ac/yr soil loss in most cases. The model has been shown to be much more accurate when the soil losses are two to three magnitudes of T. WEPS shows the range of losses in this texture group when only changing the soil.
One soil was selected from the group of soils ran to represent the Sapric Histosol soil group. Exhibit 618-16 in the National Agronomy Manual list seven texture classes plus Sapric soil materials in Wind Erodibility Group 2. They have an I factor of 134 ton/ac/yr. A random group of these textures made up of 5 LFSs, 5 LSs, 1 VFSL, and 1 LCOS were selected. The soils were selected from known wind erosion counties and weighted toward the amount of the textural class in that county. WEPS runs used in the evaluation had the same parameters, only the soils were changed to determine predicted wind erosion. All rock data was removed from the soil properties. The WEPS runs reflect the new clay adjustment and no root surfacing. The average soil loss for the group was 18.4 t/ac/yr. Gomez_Go_100_LFS was the selected soil record that calculated closest to the group average. Gomez_Go_100_LFS has a soil loss of 18.1 t/ac/yr. I have renamed the soil “Organic Soil” with an .ifc extension for the model to use. All runs using the dot ORG extension have the Gomez_Go_100_LFS data in the soil file.
Figure 1 – Soil Loss with WEG 2 soils in Grant Co. WA.
Note the range (9.8 to28 t/ac) of predicted wind erosion with the selected soils. The red bar is the Gomez LFS, renamed Organic Soil.ifc in the model.
Larry West, National Leader for Soil Survey Research & Lab, NRCS, Lincoln, NE provided summarized soil data from Michigan and from Florida. This data was used to develop two additional WEPS soil records to test against the Gomez LFS (ORG) file. A third set of Michigan data from D.L. Mokma was reviewed and considered. It was very similar to the West data. The decision was to build only two records to test using the Larry West data.
Each set of runs below compares the Gomez LFS (Organic Soil.ifc, red bar soil figure 1) to the FL and MI soils presented at the Florida close out meeting. All groups of runs except Klamath County, OR contain a standard Corn-Soybean management and a local management. All have the three soil files run with the two different managements, for a total of at least six runs per location. All graphs show the Standard run of Corn-Soybeans and a Local run with the yields set for the county. Some sites with a high rate of erosion have an additional run showing how a small management change would reduce erosion.
Each figure listed below represents a different location. Climates and field size is held constant for all runs at that location. At least two managements were run at each site. The standard run is the Corn-Soybean mulch-till, and the local run is a management that is common to the area. Some locations have an additional run that shows a solution to the excessive soil loss.
Figure 2- Grant County, WA - WEPS Organic Soil compare for Local runs and Corn-Soybean
Grant Co. WA showed very low soil loss with the standard Corn and Soybean management, and a large loss with both of the soils from Larry West’s data.
Figure 3-WEPS Organic Soil-Local & Standard Runs for Palm Beach, FL
Figure 3 seems to show a little more erosion for the average original soil than the MI soil for the corn and soybean rotation.
Figure 4-WEPS Organic Soil-Local and Standard runs for Pasquotank Co. NC
The local rotation in Pasquotank Co., NC (Figure 4), shows low rates of erosion because of No-till and Mulch-till.
Figure 5-WEPS Organic Soil-Local & Standard runs for Orange Co. NY
Figure 6-WEPS Organic Soil Local & Standard Runs for Cayuga Co. NY
Both locations in NY (Figures 5 and 6) show that the FL soil has the highest erosion on all runs. The ORG soil comes in second highest for the standard rotation and the lowest erosion with the local Green Bean run.
Figure 7-WEPS Organic Soil Local & Standard Runs for Allagan Co., MI
Note: thewinter cover crop of rye reduces the soil loss the after the potatoes from 23 t/ac/yr to 1.2 t/ac/yr for the ORG soil in Allagan County, MI.
Figure 8-WEPS Organic Soil Local & Standard Runs of Newaygo Co., MI.
Figure 9-WEPS Organic Soil Local & Standard Runs for Lapeer Co., MI
Note: The rye cover crop behind the Onions reduced erosion from 14.9 t/ac/yr to 4.5 t/ac/yr for the ORG soil in Lapeer County, MI.
Figure 10-WEPS Organic Soil Local & Standard Runs for Gratiot Co., MI
Note: The winter cover of rye for the sugarbeet rotation was planted too late (Oct 29) to be effective. Soil loss did not change.
Figure 11 WEPS Organic Soil Local Runs for Klamath Co., OR
These two managements came from Tom Golhke’s visit to the field in Klamath Co. The wheat was planted to late too provide protection after the potatoes. Erosion is controlled by planting the wheat early and adding a cover crop after the onions.
Discussion
WEPS runs were made in 10 locations using the county default locations for climate records. Nine locations included a Standard run of mulch-till management-Corn and Soybeans. The Klamath Co. location used two local managements provided by Tom Golhke, Regional Agronomist at the West NTSC. All locations have runs using the three proposed soil files: Organic Soil.ifc (ORG), Organic Soil MI.ifc (MI), and Organic Soil FL.ifc (FL).
Data for the ORG soil came from the Gomez_Go_100_LFS mentioned earlier. The two other organic soil files representing MI and FL were developed from the data sent to us from Larry West at the National Soil Survey Center in Lincoln NE.
Most of the conditions reviewed in field had a high water table or were sub-irrigated. Some of the fields had pumps and regulating structures to control the level of water in the soil profile. A flooded irrigation operation in the management file was used to simulate the extra water in the profile. This feature will be a training item to teach staff when NRCS implements the model. An alternative to the flooded irrigation feature is to use the high water table listed in the soil survey data. This will need to be investigated.
All runs were made with calibrated yields, on actual field sites with Muck soils listed on the field survey sheets, and field dimensions taken from the Web Soil Survey map generated for each site. Yields were verified using NASS data. The soil loss for all of the runs seems reasonable. All crops seem to grow well except for Radishes that needed to have the days to maturity reduced.
The Florida soil file used for the runs from the Larry West summary data shows relatively high soil losses compared to the ORG or the MI files. This is not consistent with our field observations. We thought the surface conditions were more stable in Florida than Michigan.
The group wanted more information (a baseline) before making a choice as to which record should be used. The average soil concept was proposed by Dave Lightle, Agronomist and Mike Sporcic, National Wind Erosion Specialist the first time the group met in Michigan. The concept was to use a mid range (average) for the group of soil textures listed in WEG 2 of the National Agronomy Manual. The model runs in figure 1 (mineral soil from WEG 2) provide a baseline to compare other proposed soil files.
When running the first group of soils there was a concern that the soil loss would be to low using the mid-range of WEG 2. However, after re-running the soils with the local high value crops, it looks like there will be enough soil loss generated if the T factors are adjusted to a T of 2 t/ac/yr. This means that there is soil loss when expected and practical management practices to reach T.
Using this proposed method to run the WEPS model on high organic soil seems like a reasonable approach in the short term. FL does not appear to have an extensive wind erosion problem on its muck soils. The organic soils in FL are managed with sugarcane in the rotation and have lower erosion rates. FL, MI, NY, WI, NC, SC, and OH are the significant states that will need to run the model on cropped organic soils.
Recommendations
1. Use the ORG (Organic Soil.ifc) file: Use the original average soil in WEG 2. Rename the Gomez_Go_100_LFS.ifc file and call it Organic Soil.ifc for all organic soil situations. This includes true mucks, all forms of mucks (Sapric, Hemic, and Fibric), and Histic Intergrades. It also will include any soil with an O horizon that is now incorporated in the tillage layer. Based on the runs listed above, the model will predict soil loss when we would expect it to erode and the values are high enough to indicate the need for a change in management.
2. Edit the Organic Soil.ifc file: Incorporate the bulk density and organic matter data listed in the WEPS ifc file based on the lab data that Larry West will provide by the end of the year.
3. WERU Staff, include the Organic Soil.ifc file: Have the WERU staff add the Organic Soil.ifc file to the Template subdirectory soil list in WEPS. The file will be part of the Field Office release of the model.
4. Train Staff to use the short term fix: Train staff to not use any Muck soil that has data populated, or remove the RV data from the record at the NSSC so WEPS will not use the data by mistake. Develop organic soil training exercises to be added to the WEPS 1.0 Users Manual. This will better illustrate how and when to use the short term fix. A special training effort will be made in states where organic soils occur.
5. Document: When WEPS is adopted and the revised National Agronomy Manual will contain the instructions to implement the Organic Soil file in WEPS. In addition, the instructions may also be listed in the WEPS 1.0 Users Manual, the National Soil Survey Manual, or in a National Agronomy Technical Note.
6. Evaluate: Evaluate the short term fix six months after implementing this in the field offices. The National Wind Erosion Specialist will contact states to gather and evaluate wind erosion runs made on organic soils. A report will be developed and sent to the National Agronomist evaluating the short term fix.