Fashioning Amendments and Seeding Techniques to Rebuild Essential Genera of Endophytes and other soil Nutrient mobilizers (FAST REGEN)

This 2-year, 2-farm research project used a mix of three cover crops – winter wheat, Austrian peas, and winter camelina – and three supporting concepts to try to replace herbicide and chemical fertilizer in commercial field corn production. These three crops were expected to provide distinct and complementary soil services that would improve yields of subsequent crops, such as the following: Legumes add nitrogen to the soil, grasses improve soil structure, and brassicas add biomass and pest controlling chemicals. Camelina is a wispy plant, especially for a brassica, but it is extremely cold tolerant. 

Concept 1: Anti-herbicide compost

To remove herbicide residues that might impede cover crops, in year 1 we attempted to cultivate microbial metabolizers of 2,4-D, atrazine, glufosinate, and glyphosate by inoculating compost with these herbicides. We mixed approximately equal volumes of deciduous leaves and spent coffee grounds in a 21 cubic foot ComposTumbler®, inoculated with herbicide treated soil and standard compost. Herbicide exposure began on June 15, 2023 at 6.25% of standard application rates for 10 cubic feet of 6-inch-deep topsoil (20 square feet of surface area) after several weeks of composting, and doubled every 7-14 days for 6 weeks, reaching 200% on August 14, 2023. A ~2 cubic foot tumbler was maintained in parallel with similar feedstocks but without herbicide exposure. 

The 100% application rates were

Of the herbicides applied, atrazine was by far the most difficult to dose accurately. The amount of concentrated atrazine needed to inoculate the compost tumbler is only 0.86 milliliters at the 200% dose. 100% atrazine is very viscous and hard to lift into or expel from a pipette. The doses actually applied could have been in the 0.5 to 1.5 mL range. I did not mix a diluted solution because I was unsure of the stability of diluted atrazine over the storage period. The other herbicides are much less viscous and are easy to measure accurately. All herbicides were mixed together in a gallon of water which was poured across the surface of the compost in the 21 cf tumbler. The tumbler was rotated 360 degrees after each herbicide application and occasionally between applications. 

Anti-herbicide leachate was made by mixing about equal volumes of this compost with rain water, then pouring the mixture through a large cold brew coffee filter. The resultant solution did not clog sprayers, e.g. a DeWalt backpack sprayer. Anti-herbicide leachate could be used alone or mixed with similarly obtained symbiotic compost leachate. Leachates were either sprayed into a cement mixer as wetting agents to make seed pellets and seed bombs, or applied directly to plots by spraying. 

On 10/24/2023 samples of AHC and the symbiotic composts were delivered to a collaborating university before the seed pellet production began. Our intent was to analyze the capacity of the compost to metabolize herbicides, indirectly supporting its inclusion in seed pellets to clear herbicide residues ahead of a growing seedling. Unfortunately, our collaborator at that institution took a position at another institution, the freezer containing the original compost samples was reorganized, and the samples were determined to be irretrievably lost in August 2025. Meanwhile, as warnings accumulated that our original university affiliation might not be fruitful, we began complimentary field studies. In light of the lower importance of the post-emergent herbicides and difficulty in analyzing them, we focused on atrazine, the only pre-emergent herbicide in the trial. 

In August 2025 we confirmed that our atrazine stock remained herbicidal, added about 50 pounds of spent coffee grounds to rejuvenate the original tumbler, and re-inoculated that tumbler with two doses of atrazine 8 days apart. We estimated the volume of compost in the tumbler using geometry, and its mass based on measured density. On 8/19/2025 we treated ~205  L (~107 Kg) of anti-herbicidal compost in the original large anti-herbicide compost tumbler with about 0.2 mL of 43% atrazine in ~3.75 L (1 gallon) of rain water, trying to achieve a 1 part per million atrazine dose. On 8/27 we added an additional 1 mL of 43% atrazine which would raise the dose to about 6 PPM. 

The control compost was treated similarly in its tumbler, but without atrazine. The tumblers incubated for 60 days at an estimated average temperature of 67°F, with the high exceeding 70°F on 54 of the first 55 days, 80°F on 38 days, and 90°F on 8 days. On 10/26/2025 we froze 150mL samples for LC-MS-MS analysis at Dr. Chung Ho Lin's laboratory at the University of Missouri in Columbia, MO. These were received on 10/28 and results returned on 12/10/2025. The anti-herbicidal compost atrazine level was 1,067 ug/kg (parts per billion), and the control level was 33.5 PPB. If the initial concentration of atrazine was 6 PPM (6,000 PPB), then its half life under these conditions was about 24 days. Although this is less than half of the commonly estimated half life of 60-75 days in soil, the shortest reported atrazine half lives are even shorter, less than 2 weeks. Also, several more half lives would need to pass before the anti-herbicidal compost could be safely incorporated into a seed pellet, especially for a brassica like camelina or radishes. Although the sample of anti-herbicide compost used in the seed pellets was lost, the original final inoculation and use schedule was similar to the the schedule just described, so the anti-herbicidal compost could have been more toxic than protective in the seed pellets, even after dilution with other materials. 

Disappointed by this result, and to resolve some of the remaining uncertainty about the anti-herbicide compost efficacy against atrazine, I began a controlled decomposition trial. On 12/16/2025 (about 2 more half lives after the sampling on 10/26/2025) I took 3L of control and 3L of anti-herbicide composts from the tumblers, each of which had a dry weight of about 778 grams. I froze a ~100 g dry weight baseline sample of each compost. To each compost in a separate galvanized metal bucket, I added 15.8 mL of a 1:1000 dilution of 43% atrazine in 250 mL of rain water, seeking to achieve a dose of 10 mg atrazine/Kg dry compost (10 PPM). I thoroughly mixed the composts and atrazine solutions and took a second pair of ~100 g dry weight samples for immediate freezing. The buckets were then incubated with thermostatically controlled heating mats in an insulated aquarium in a loosely temperature controlled greenhouse until 1/12/2026 (27 days), when a third pair of ~100 g dry weight samples was collected and frozen. Compost temperatures were documented between 15 and 27°C (59-80°F) during the incubation. The control compost was consistently 1°C warmer than the AHC. All three pairs of samples were sent to Dr. Lin's laboratory, where they await analysis as of 1/31/2026. When the atrazine assays are completed we will amend this document to report the atrazine half life in AHC versus control compost. 

2023 subsidy of herbicides versus AHC

In a 2023 sub-study parallel to the main trial, I divided a 20' x 60' conventionally managed strip into four 15' sections, and on 10/21 sprayed each with a full dose of one of the herbicides, as illustrated here (Field study of herbicides in 2023). On 10/31 I planted eight rows of seeds, each spanning all four herbicide conditions. The eight rows were two of bare seeds, two of bare seeds with anti-herbicide leachate, two of calcium bentonite seed pellets with anti-herbicide leachate, and two of sodium bentonite (Volclay) pellets with anti-herbicide leachate. 

On 11/13 Microbiometer assays were done to evaluate soil microbe activity, measured in µg C / g soil, with a percentage of fungal versus bacterial activity (these always sum to 100%). The herbicide treated sections ranged from 178 to 211 µg C/g, with the lowest activity in the section treated with glufosinate (178, 15% fungal). For comparison, the comparison plot treated with glyphosate but fertilized with compost and trace elements had total activity of 290 (34% fungal), the comparison plot fertilized with NPK but not treated with herbicides had activity of 223 (29% F), and the comparison plot fertilized with compost and trace elements without herbicides had activity of 366 (41% F). These results are consistent with the expectation that herbicide treatment impairs the microbiome, particularly fungi. 

By 11/17 only the Volclay rows had produced a minimal number of seedlings. No peas were found in the 2, 4-D section. Seedlings emerged at similar densities in the other 6 rows, with wheat being the most successful plant by far, and camelina next. The success of many Phuel seed pellets that incorporated anti-herbicidal compost suggests that the residual level of atrazine in the AHC was not herbicidal. By March 2024 no cover crop plants remained in the atrazine section, regardless of attempted AHC protection, while cover crop plants remained in the other three sections.  We concluded that residual atrazine in fields was the most prominent herbicide problem to address. 

2024 substudy of residual atrazine concentrations

On 9/3/2024 I began a sub-study to evaluate the interaction of atrazine and seed pellets. I divided the same 20' x 60' conventionally managed, herbicide treated strip into 9 equal sections, and sprayed these with doses of atrazine simulating 0, 1, 2, 3, 4, or 5 half lives of atrazine (Field study of atrazine in 2024). The non-random treatment pattern placed higher doses of atrazine downhill from larger doses to preserve an atrazine gradient in the event that heavy rainfall carried some herbicide downhill.

On 9/6 the AHC from the previous year was rejuvenated with spent coffee grounds and hardwood sawdust, and re-innoculated with 0.75 g of atrazine, or about 750 parts per billion in the estimated 100 kg of compost. The compost was kept moist and tumbled weekly until mid November, when it was used to make new seed pellets with wheat, pea, clover, camelina, and radish seeds. The very small camelina and clover seeds were combined in seed bombs rather than single seed pellets. 

On 11/18 weeds were qualitatively assessed in the western third of the plot, where the atrazine dose range was maximal. The section with a 100% atrazine dose (0 half lives on 9/3) grew wild garlic (Allium vineale) almost exclusively. The sections with 13% and 1.5% doses (3 and 6 half lives, respectively) were qualitatively similar to the remaining sections, and were supporting the following species, as identified by the Picture This app, in order of abundance: [1] Wild garlic; [2 & 3] Carolina crane's bill (Geranium carolinianum) & Henbit (Lamium amplexicaule); [4] Wild carrot (Daucus carota); [5] White clover [Trifolium repens]; [6, 7, & 8] Canadian goldenrod (Solidago canadensis), Groundsel tree (Baccharis halimifolia), & Virginia plantain (Plantar virginica).

A sample of the anti-herbicidal compost was delivered to a collaborating university before the seed pellet production began. Our intent was to analyze the capacity of the compost to metabolize herbicides, indirectly supporting its inclusion in seed pellets to clear herbicide residues ahead of a growing seedling. Unfortunately, our collaborator at that institution took a position at another institution, the freezer containing the original compost samples was reorganized, and the samples were lost. In light of the lower impact of the post-emergent herbicides, we focused on atrazine, the only pre-emergent herbicide in the trial. We confirmed that our atrazine stock remained herbicidal, added about 50 pounds of spent coffee grounds to rejuvenate the original tumbler, and re-inoculated that tumbler with two doses of atrazine 8 days apart. We estimated the volume of compost in the tumbler using geometry, and its mass based on measured density. On 8/19/2025 we treated ~205  L (~107 Kg) of anti-herbicidal compost in the original large anti-herbicide compost tumbler with about 0.2 mL of 43% atrazine in ~3.75 L (1 gallon) of rain water, trying to achieve a 1 part per million atrazine dose. On 8/27 we added an additional 1 mL of 43% atrazine which would raise the dose to about 6 PPM. 

The control compost was treated similarly in its tumbler, but without atrazine. The tumblers incubated for 60 days at an estimated average temperature of 67°F, with the high exceeding 70°F on 54 of the first 55 days, 80°F on 38 days, and 90°F on 8 days. On 10/26/2025 we froze 150mL samples for LC-MS-MS analysis at Dr. Chung Ho Lin's laboratory at the University of Missouri in Columbia, MO. These were received on 10/28 and results returned on 12/10/2025. The anti-herbicidal compost atrazine level was 1,067 ug/kg (parts per billion), and the control level was 33.5 PPB. If the initial concentration of atrazine was 6 PPM (6,000 PPB), then its half life under these conditions was about 24 days. Although this is less than half of the commonly estimated half life of 60-75 days in soil, the shortest reported atrazine half lives are even shorter, less than 2 weeks. Also, several more half lives would need to pass before the anti-herbicidal compost could be safely incorporated into a seed pellet, especially for a brassica. Although the sample of anti-herbicide compost used in the seed pellets was lost, the original inoculation and use schedule was similar to the the schedule just described, so the anti-herbicidal compost could have been more toxic than protective in the seed pellets, even after dilution with other materials. 

Disappointed by this result, and to resolve some of the remaining uncertainty about the anti-herbicide compost efficacy against atrazine, I began a tightly controlled decomposition trial. On 12/16/2025 (about 2 more half lives after the sampling on 10/26/2025) I took 3L of control and 3L of anti-herbicide composts from the tumblers, each of which had a dry weight of about 778 grams. I froze a ~100 g dry weight baseline sample of each compost. To each compost in a separate galvanized metal bucket, I added 15.8 mL of a 1:1000 dilution of 43% atrazine in 250 mL of rain water, seeking to achieve a dose of 10 mg atrazine/Kg dry compost (10 PPM). I thoroughly mixed the composts and atrazine solutions and took a second pair of ~100 g dry weight samples for immediate freezing. The buckets were then incubated with thermostatically controlled heating mats in an insulated aquarium in a loosely temperature controlled greenhouse until 1/12/2026 (27 days), when a third pair of ~100 g dry weight samples was collected and frozen. Compost temperatures were documented between 15 and 27°C (59-80°F) during the incubation.  All three pairs of samples were sent to Dr. Lin's laboratory, where they await analysis as of 1/31/2026. When the atrazine assays are completed we will report the atrazine half life in trained versus naive compost. 

Concept 2: Epigenetic adaptation

We planted commercially purchased winter wheat, Austrian pea, and winter camelina seeds in monocultures on regeneratively managed, herbicide-free 400 square foot plots in the fall of 2022 for both seeds and bacterial inoculants for compost piles. The seeds collected from these plants in the summer of 2023 should have had locally relevant genetic traits and epigenetic adaptations (GNE19-028). The wheat plants had problems in part due to aging seed stock and subsequent weed competition. Peas and camelina yielded sufficient seeds for the main trial. Wheat seeds for the main trial were collected from a large nearby field that Mr. Altoff had planted and harvested with a combine. 

Five gallons of wheat seeds for the experiment were collected by combine from a field that Mr. Altoff managed conventionally. Pea and camellia seeds were collected by hand from plots that Dr. Sumner managed. Camelina pods began to shatter before being harvested, self seeding the plot where they grew and eventually generating a second harvest. In spite of the shattering, we collected 114 grams of seeds from the first crop in a 20'x20' plot, and estimated 540 viable-looking seeds per gram, for a presumed viable seed collection of over 66,000 camelina seeds. Similarly, we found that 100 mL of dried peas weighed 92 grams and contained about 470 viable looking seeds, and collected about 23,000 seeds from a 400 sf plot. We had originally sought a minimum of 3 seeds of each species per square foot in 3 replicates of 6 conditions on 400 sf plots, for a requirement of 21,600 seeds for the original experiment. Peas were obviously the most likely to fall short if more plots were added. 

We did not directly evaluate seed adaptation. 

Concept 3: Symbiotic composts

We composted plant tissues (e.g., panicles, chaff, non-viable seeds, root samples) and soil from the monocultures to obtain crop-specific endophytic bacteria and other beneficial microbes. We inoculated leaf-and-coffee compost in crop-specific 18 cubic foot Rubbermaid® compost bins. Following a refinement of this concept, we are now growing mixed and individual cover crops in compost piles. Plants grew reasonably well in a compost pile in full sun, which had a relatively low level of microbial activity. Plants grew very poorly in a pile in partial shade, although microbial activity was significantly greater.  

Biochar

We produced biochar from brush piles using Ring-of-Fire flame-capped kilns, and sifted it through hardware cloth and wire mesh to obtain different grades, including dust-like ultrafine and coarse (¼ to ¾ inch) grades. The pH of biochars that we produced in flame capped kilns is consistently higher than 9, and can be 10-11, even after extensive rinsing with rain water. This seems high enough to damage symbiotic microbes, although we have not had the means to confirm this fear. The pH falls to 7+/-1  with the addition of at least 0.4% by volume of concentrated sulfuric acid, e.g. 1 ml H2SO4/236 ml (1 cup) of biochar. Finely ground biochar, having a layered surface area, may require larger amounts of sulfuric acid to reach a neutral pH. 

In general we can bring biochar close to neutral pH by adding 70-80 mL of H2SO4 diluted in 200 mL of water to 5 gallons of 1 inch minus biochar. We made 300 gallons of biochar for an additional field trial arm, placed in buffer zones between seed pellet plots.

Seed pellets

In the fall of year 1 we manufactured separate seed pellets (a single seed with a protective and/or nourishing coating) with winter wheat and peas as nuclei. Camelina seeds are too small to work as nuclei of seed pellets but were added as an outer layer to pea or wheat seed pellets, creating a seed bomb (a ball with a lot of seeds).  We tested two cement mixers and several types of bentonite clay. We settled on the Yardmax 1.6 cf mixer with paddles removed. We did not find a dominant clay after trying Wyoming bentonite (sodium), Volclay (calcium bentonite), Indian Healing Clay (sodium bentonite), Seed Ball Phuel mix and Red Clay Powder for Seed Balls and Seed Bombs (both at Seed-balls.com at Amazon.com), and unscented cat litter. We made most of the seed pellets with Phuel, red clay, and Wyoming bentonite. 

Peas consistently swelled in the center of a moist seed pellet as it dried, very frequently cracking the pellet and exposing the pea and leaving it vulnerable to  predation. We tested a few ways to wrap the peas in a layer to prevent swelling, e.g. a light coating of plant oil or a sticky sacrificial hydrophilic layer like molasses. These did not prevent swelling. 

FAST REGEN

We began a block randomized experiment on 0.4 acres of a 2-acre field managed by Mr. Altoff to compare 8 treatments’ effects on soil chemistry, herbicide residues, and microbial activity; cover crop mass; and field corn yield and elemental composition in year 2. Due to significant weather problems, we planted multiple additional plots at Sumner's Farm to evaluate the individual concepts at smaller scale, and performed some experiments using seed starting trays and pots.

Plots in the 2-acre field were 20-foot squares (400 sf each), in triplicate for each treatment (8*3 =24 blocks total). All blocks receive conventional herbicides and fertilizers in the spring of 2023. A soybean crop was planted in the summer of 2023 but failed completely due to drought, leaving the field to become overgrown with crab grass and Palmer amaranth. These weeds went to seed before they were cut and sprayed with glyphosate in late October.  The originally planned cover crop seeding was completed on October 23 in anticipation of rain that night, but only a minimal rainfall occurred, followed by additional weeks of drought.  There were originally 6 seeding conditions repeated in 3 blocks, but space allowed for 2 more conditions in each block. We added one condition with 100 gallons of near finished compost plus trace elements shallow tilled (4" depth) into 400 sf, and a second condition with 100 gallons of acid-neutralized biochar plus fish emulsion, compost leachate, and trace elements shallow tilled into 400 sf.  Uncoated cover crop seeds were scattered into the new plots and raked on October 31, just ahead of the first freezing weather of the season.  Composted pig manure was spread over a 0.2 acre strip of the 2-acre field in early December. 

Treatments

AHBC=Anti-Herbicide BioChar; AHCL=Anti-Herbicide Compost Leachate; GSC = Generic Sifted Compost + Mycorrhizal inoculant; SCL=Symbiotic Compost Leachates (mixed leachate from all cover crop composts).

Trace element additions were prompted by a soil sample showing 1.82 ppm copper, <0.02 ppm molybdenum, and 0.42 ppm boron. 

Wheat and pea seeds germinated in November did not do well, and did not cover the soil as hoped. Camellia was hard to distinguish from small dicot weeds. We terminated cover crops with a 4' wide crimper roller on May 11, 2024, shortly prior to corn planting in year 2. Field corn planting and harvest proceeded as usual. As expected from the poor cover crop coverage, weeds were very plentiful and were not shaded by the cover crop or killed by the crimper. Wheat was terminated, but rebounding weeds competed with and seriously impaired the corn plants, which ultimately yielded nothing. Some pea plants survived the crimping. 

Soil microbiome activity was assessed in February 2024 in the 2-acre field. Only three representative samples of the FAST REGEN plots were measured because of the uniformly disappointing cover crop establishment. N=3 for all rows in the table. Microbiometer assay results were:

FAST REGEN Small scale backup study

Backup plots on Sumner's Farm were planted and watered in late October as the drought and weed crises unfolded on the 2-acre field.  These backup plots were:

1. No prior treatment, cover crop
2. Pampered with compost, weed removal, watering, and straw cover for the cover crop
3. Cover crop and NPK
4. Glyphosate and NPK
5. Compost for 3 years
6. Compost and trace elements for 3 years

Backup plots 5 and 6 grew healthy stands of weeds, and were never seeded with cover crops, as they had already been treated for several years with a rich compost made from coffee grounds, dry leaves, and wood chips. The plots were in line with the backup plots 1-4, and were therefore simply planted with corn at the same time as the intended backup plots. They did not get cover crops or extra fertilizer or glyphosate treatment. Subsequent photographs document that these two plots completely outperformed the backup plots and the seed pellet plots in the 2-acre field, as well as the conventional plots in the 2-acre field and other nearby fields. Eventually other observers opined that the corn in backup plots 5 and 6 was 6 weeks ahead of conventional corn. Some objective measures were as follows:

Before crop samples were collected from these plots, in late July or early August, every plant in these two plots was bent to the ground, and practically every kernel on every ear of corn was eaten, with racoon and deer tracks evident. Nearby, corn stalks in a conventionally managed, completely unfenced deer plot remained standing through season's end, with many kernels left on most ears. Having seen this before, we have come to believe that vertebrate dietary preferences reflect crop quality. 

We monitored plots as well as compost piles using microBiometer assays. We used an AtLeaf meter to obtain a nondestructive chlorophyll measure on corn leaves, and optical Brix measures on liquid expressed from corn leaves crushed in a garlic press. These measures were taken from the second lowest leaf on a corn plant, or a nearby leaf if the target leaf was compromised in some way. AtLeaf measures were taken within 3 inches of the stalk.