Using Living Aisles and No-Till Planting Strips to Mitigate the Impacts of Intense Rain Events on Organic Vegetable Farms

Cover Crops – Living Aisles

Visual assessment of the clover/oat aisles and oat/pea vegetable planting strips in October 2018 was a bit better than expected given that planting was delayed due to early September rains.  Population counts for the white clover averaged from 54 (Blue Moon), to 65 (Equinox), to 69 (Two Onion) plants per square foot 14 days after seeding.  These counts were less than desired based on our previous work with living aisles in 2015 and 2016.  In those trials the average population density of the white clover was 120 plants per square foot. 

Assessment of the clover aisles on 4/19/19 showed that all the clover at Two Onion (TO) had died over the winter and that the population densities at Equinox (EQ) and Blue Moon (BM) were not adequate to exclude weeds.  The aisles at TO were reseeded at the full original seeding rates for both the clover and oats on the day of the visual inspection and filled in very nicely though the spring and summer.  The aisles at EQ and BM were reseeded with only clover and at half the rate of the original fall planting, just to fill in gaps.  These aisles also filled in nicely through the 2019 season.

In 2020, the aisles were maintained with occasional mowing.  The farmers at BM at TO reported that the clover flowered and looked pretty good overall.  No other action was taken in the plots in 2020, due to the pause caused by COVID-19.

Visual assessment of the clover aisles on 4/8/21 showed that the clover was coming in pretty well at TO and that there were not many weeds apart from the field edges where quack grass was infiltrating.  At BM the clover populations had diminished and weed populations had grown.  At EQ the clover was overtaken by weeds so completely as to no longer be able to distinguish where the aisles had been, and the plot was abandoned. 

Cover Crops – Planting Strips

Fall dry biomass of the oats and peas in 2018 was also less than desired.  This winter-sensitive cover crop was planted to form an in situ mulch for two of the no-till planting strip treatments in the spring of 2019.  The more biomass in the fall, the more residue that should be in place as a mulch in the spring to exclude weeds.  Though we did not expect to produce enough biomass to exclude weeds for the full 2019 growing season, we did hope to at least exclude weeds through pepper planting in late May.  The biomass as detailed in the numbers below, was far less than our goal of 5,000 pounds per acre, and indeed was not enough to exclude weeds through pepper planting.

Dry Biomass in lbs/acre on 10/30/18

Oats and peas planted in the planting strips in fall 2019 winter killed as expected, but were not planted to vegetables or replanted to oats and peas in fall 2020, due to the pause caused by COVID-19.  As a result, there were no measurements taken on the planting strip cover crops in 2020 or 2021.  

Weed Control

Individual weed population counts taken 14 and 49 days after cover crop seeding in 2018 ranged from 0 to 19 weeds per square foot across the three farms and averaged 1.9 weeds per square foot.  The vast majority of these weeds were annuals which did not produce seed before dying over the winter.  Annual weeds were suppressed by clover and mowing in the aisles and with occultation, mulch, or tillage in the planting strips in 2019, as expected.  The few perennials in the mix were more problematic and were managed by mowing the aisles and hand-pulling the planting strips.

The pause in 2020 resulted in many more annual and perennial weeds infiltrating the plots.  Weed counts taken in the pepper planting strips on 6/22/21 at TO and on 6/24/22 at BM ranged from 12 to 112 weeds per square foot, with and average of 32 in the tilled treatment (TO only), 57 in the occultation treatment, 46 in the control, and 0 where the landscape fabric mulch excluded weeds in the mulched treatment (except in planting holes).  More importantly, as would be expected after the pause in 2020, the perennial weeds were much more prevalent and harder to control with regular weeding.  At both BM and TO quack grass came in from the edges of the overall plot.  All the control plots, and one of each of the treatment plots at TO were on the edge and quack grass infiltration was so severe as to severely stunt the pepper plants in particular.

Broccoli Yield

Every broccoli plant in every treatment and control plot at every farm buttoned in 2019 and in 2021.  The most likely cause of buttoning in broccoli is stress.

In 2019, though the seedlings looked healthy when planted, they were in their cells for about a week longer than planned, which may have caused stress.  They were covered with row cover at planting to prevent flea beetle damage and preserve warmth on the cold spring nights, as is common practice at both EQ and BM, though not at TO where flea beetle pressure is less.  Row cover was removed after three weeks for weeding, at which time it was already clear that all the broccoli would button.

In 2021, the seedlings were healthy when planted and did not seem at all stressed.  At BM, the plants were covered with row cover immediately after planting, as in 2019.  When the plants were uncovered two weeks later, they appeared wind abraded, purple, and obviously stressed.  At TO, however, the plants were not covered with row cover, in accordance with their typical practice for brassicas.  If the row cover had been the cause of stress, the broccoli at TO should have headed as normal.  The broccoli at both BM and TO did button, however.

In both 2019 and 2021, broccoli was planted in late April.  Though the weather in Wisconsin is still unsettled at that time, it is common for farms in the area to plant a first or second succession of broccoli late in the month.  BM, in fact, did plant broccoli on the same day as the trial planting in 2021, and that planting did not button, though it was treated in the same way as the control in this trial.  In short, there were no differences between the treatments and control in terms of broccoli buttoning, and  the exact reason for the buttoning in this trial is unknown.

Yield on the buttoned broccoli was recorded both years and ranged from an average of 0 to 0.22 lbs/head.  As the buttoning was the major contributor to yield, not the treatments, the data was not analyzed.

Pepper Yield

Marketable green peppers were harvested weekly, sorted into first and second quality fruits, and weighed.  Non-marketable peppers were left in the field.  First quality fruits were mostly blemish free, with one occasional small blemish on a large fruit.  Second quality fruits were small in size or had larger healed blemishes.  Non-marketable fruits were very small, diseased, or had open wounds.  

In 2019, visual observation of the plots and the raw yield data both indicate that the control outperformed all three no-till treatments with living aisles.  Generally it appeared that the peppers were both larger and more numerous in the control plots.  Yield data for 2019 is still being analyzed.  When analysis is complete, we expect it to confirm these observations. 

In 2021, numerous factors contributed to yield outside of the treatments.  At TO quack grass infiltrated from the edges of the trial area and impacted every edge plot.  Six of the 12 total pepper plots were on the edge.  The three control plots, in particular, were taken over by quack grass that became impossible to terminate with hand weeding and clearly stunted the plants and reduced yield (average 0.12 lbs/plant).  Even the one mulch treatment plot on the edge had enough quack root pressure to reduce the average yield by half (1.91 lbs/plant) as compared to the other two mulch plots (3.78 and 4.07 lbs/plant).

At BM in 2021, a late frost killed all but 1 plant in the mulch plots, and all but 5 plants in the occlusion plots.  By comparison, 40 plants survived the frost in the control plot.  The farmers speculated that more heat escaped the ground in the tilled control than in the no-till plots, thus protecting more plants in the control.  Though all frosted plants were replaced from the same batch of pepper seedlings, the replacements were transplanted seven days later than the originals.  This later transplant date may have contributed to a difference in their performance and by extension a difference in the performance of the mulch (average 0.59 lbs/plant) and occultation (average 0.14 lbs/plant) treatments as compared to the control (average 1.69 lbs/plant) where many fewer plants were replaced.

Because these extenuating factors made it impossible to attribute yield differences to the treatments, pepper yield from 2021 was not analyzed.

Labor Overall

Management time for each treatment was tracked in minutes for planting and maintenance (weeding, picking up mulch, etc.). Because harvest labor is primarily dependent on yield and is not affected by the treatment, it was not tracked and is not included in the labor totals.  The chart below shows the total average labor in minutes for a 100’ bed by crop, treatment, and activity. 

Overall Average Labor per 100' Bed in Minutes

Broccoli Labor

Planting - The real differences in labor time show up in planting.  The occultation and mulch treatments took longer to plant because individual holes had to be dug by hand into untilled soil for each individual seedling.  By contract, in the control and tilled treatment, seedlings were tucked into the soft soil with minimal effort.  The mulched treatment took even longer to plant because of the extra time needed to secure the mulch and finagle each seedling into the inflexible hole in the landscape fabric.

Maintenance - In the mulch treatment, weeding time was less than the other treatments or control, but the time needed to remove the landscape fabric after harvest more than made up for that savings in maintenance time overall.  Given that spring broccoli is a fairly short lived crop in the field, requiring little time in weeding, the weed exclusion benefit of the mulch did not pay off in maintenance time overall.

Pepper Labor

Planting - As with broccoli, it took more time to plant peppers into the occultation and mulch treatments given the extra effort needed to dig individual seedling holes into untilled ground.  Generally less planting time needed for peppers as compared to broccoli can be attributed to wider spacing between plants and thus fewer seedlings planted per 100’ bed.

Maintenance - Given the longer pepper season, there is more nuance to the maintenance time discussion, as compared to broccoli.  Not surprisingly, the mulch treatment required less weeding time.  Further, with the longer season and thus higher weeding requirement in the other treatments, the time invested in putting down and picking up the mulch was more than covered by time saved in weeding.  More surprisingly, the time needed to mow the living aisles in the mulch treatment also was not enough to top the maintenance time in the clean cultivated control without a living aisle. 

Discussion

Climate change predictions for the Midwest include more intense rain events interspersed by periods of drought.  The farms in this trial experienced both conditions over the course of this project, emphasizing the need for research into methods for absorbing excess rainwater and for making it available to crops during dry periods.  Because we already know that living plant material and undisturbed soil absorbs water and reduces erosion during rain events, the basic goal of this research was to test if yields could be maintained with living aisles and no-till planting methods.

Unfortunately, every broccoli plant at every farm buttoned in both years of the trail, making the broccoli yield data useless.

Further, the project was paused in 2020 due to travel and other restrictions aimed at containing COVID-19.  During the pause, the trial plots were mowed, but otherwise left untouched, leaving them vulnerable to invasion by perennial weeds from the field edges. Perennial weeds are a common problem in organic no-till and reduced tillage systems and in 2021, weed pressure at EQ was so strong as to make the plot unusable for the trial.  Weed pressure from the edges of the plots at TO severely compromises six of the 12 plots at that farm.  Other unforeseen issues, including a late frost and broken tiller at BM, severely reduced the value of the pepper data there.

In the end, the 2019 pepper yield data was the only useful yield data.  It showed that the tilled control with no living aisles outperformed the three no-till treatments.  We know that competition for water and nutrients, especially during the critical weed-free period, can reduce yields in vegetable crops.  Presumably, the living clover aisles presented enough competition to cause yield drag.

If there had been a savings in labor with any of the three treatments in comparison to the control, then it might have been possible to financially balance the yield drag with labor savings.  But the control was also the most time efficient system.

Though this project did not show the living aisles treatments to be the same or better than the conventionally tilled control, the farmers agreed that the benefits of living aisles and reduced tillage in relation to water infiltration and reduced erosion warrant further research.  In anticipation of future trial designs, this project offers the following insights:

• For a longer season crop like peppers, using landscape fabric mulch does appear to be worth the effort of application and removal in reduced weeding time, but not for a shorter season crop like spring broccoli.
• A 2- or 3-week occultation of a planting strip requires additional labor in laying down and picking up silage plastic, but does not offer a payback in longer-term weed control, especially when perennial weeds like quack grass can infiltrate from the field edge.
• A tilled planting strip requires no more time to plant than a conventionally tilled control, but does require more time to maintain through mowing of the living aisles.  If yield can be equalized between the two, the additional time spent in maintenance might be worth the return in reduced erosion and increased rain water absorption.
• One consideration for equalizing yield between living aisle systems and conventionally tilled systems could be the timing and height of mowing of the clover aisles.  Would mowing more frequently and shorter reduce competition with vegetable crops?

Finally, this project reinforced two of the most basic understandings in agriculture.  1) Competition from weeds or other living plants can reduce crop yields.  And 2) tillage can be an effective way of controlling weeds.  These understandings are part of why positive results in cover crop based no-till are so elusive in organic systems where herbicides cannot be used for cover crop termination or weed control.