Final Report for ONE08-082
Project Information
An on-farm research trial with two farm collaborators was undertaken to look at weed control in organic corn production through narrow row spacing compared to conventional row spacing (15” vs 30”) and use of mid-season cover crops. The effect of population thinning was measured at 8 and 11% for narrow rows and 9 – 18% for wide rows. The effect of cover crops at both sites was insignificant. At site 2 they didn’t establish and at site 1 they established in the wide rows that had been cultivated but growth was minimal since they were suppressed by the shade of the fully developed corn plants. Experience from this trial suggests that cover crops need to be established the season prior and suppressed before the current growing season’s crop establishment. There were row spacing and weed competition challenges at both sites that need to be considered when comparing yield results. At Site 1 narrow rows yielded 40.6 bu/ac and wide rows 123 bu/Ac dry shell corn. At Site 2 corn was harvested as silage and yielded 7 tons/ac. There was no statistically significant yield difference between 15” and 30” rows.
Introduction:
Weed control is one of the greatest challenges in organic row crop production. In New York State, no-till production for row crops has not been widely adopted. Since fungicide and insecticide seed treatments are prohibited under organic standards, improving soil conditions to minimize the potential for seedling rots and insect depredation is accomplished by tilling to loosen and warm the soil in the spring. The growing season in upstate NY is fairly short and soil temperatures are cool through May. Typically, mechanical weed control will include a combination of blind cultivation followed by row cultivation. Cultivation passes typically begin before the corn has emerged with flex-tine harrows or rotary hoes to control young weeds in the white thread stage. Once the corn reaches about 6 inches in height row cultivation is used. Cultivation is time intensive. The number of necessary cultivation passes will vary with growing conditions. In some years excess moisture will delay or prohibit necessary and timely cultivation passes.
There were three main objectives of this study. The first was to assess the impact of narrow row spacing (15 inch rows vs 30 inch rows) on weed control. Could narrow row spacing eliminate the need for cultivation after early season tine weed harrowing? The second was to measure population thinning from tine weeding and the third was to assess the weed suppressive effects of cover crops established at the time of the last cultivation. Legume and grass cover crops were inter-seeded for observation and comparison.
Cooperators
Research
This study had two field sites with two cooperating farmers. Twin Oaks will be referred to as site 1 and Bundy Creek as site 2. Research plots were established on first year corn fields on the two cooperator farms using a randomized complete block design. The primary treatment was row spacing which ran the length of the field thereby limiting randomization of treatments. The treatment,row spacing, was replicated four times across the width of the field. Four blocks were established perpendicular to the planted rows. Row spacing; either 30 inch or 15 inch was the main treatment and cover crop was a sub-treatment. Plant population per acre were kept consistent for both row spacings. Cover crops were seeded for comparative evaluation as a sub-plot treatment. There were four variations of cover crops; annual ryegrass, red clover (grazing type), berseem clover (an annual) and a mix of annual ryegrass and red clover. Cover crops were broadcast seeded after the final cultivation. Treatment effects were compared by measuring yields. All plots were tine weeded at least once. The 30” plots were also cultivated with a row crop cultivator. Population counts were taken twice each site. Cover crops were seeded by hand broadcasting from July 11- 14.
Site 1 (Twin Oaks). The field was soil sampled for a nutrient analysis which was done by the Cornell Nutrient Analysis Lab. Both phosphorus and potassium levels tested in the high range. Fifteen tons of liquid manure was applied prior to planting. The field was disked and culti-packed before planting. The field was planted on May 21, 2008 with a 6 row planter. The corn variety American Organic B38 was planted at 36,000 seeds/ac in 30” rows and 19,000 seeds/ac in 15” rows. The field was tine weeded 3 times about 1 week apart on May 24, May 30 and June 7. The wide rows were cultivated the following week on June 12 with a final cultivation on June 28.
Site 2 (Murray farm). The field was soil sampled for a nutrient analysis which was done by the Cornell Nutrient Analysis Lab. Both phosphorus and potassium levels tested in the medium range. Five tons of semi-solid manure was applied prior to planting. The field was moldboard plowed and disked before planting. The field was planted on May 25, 2008 with a 6 row planter. The corn variety Masters Choice 470 was planted at 34,000 seeds/ac in 30”rows and 17,000 seeds/ac in 15” rows. The field was tine weeded 1 time on June 7. Rain did not allow subsequent passes before the corn grew too tall. Rainfall also limited cultivation of the 30” rows to one time on July 2.
Harvest
Site 1. Corn grain was hand harvested October 8-13 to measure yield. Two rows, forty foot in length were hand harvested from each sub plot, with and without covercrop. Ears were weighed. A 10 ear sub sample from each sub plot was randomly selected and oven dried for moisture testing. The corn was shelled from the cob to estimate percent shell corn. Yield is reported in bu/ac at a standardized 15.5% moisture.
Site 2. The weed growth was too intense to allow hand harvest of the plots at site 2. These plots were machine harvested October 5 as silage and the loads were weighed. Yields were standardized to 65% moisture.
All plots were tine weeded at least once. The 30” plots were also cultivated with a row crop cultivator.
Population counts were taken twice at both sites. At site 1, population counts were taken after the first and 2nd tine weeding. At site 2 there was an initial population count and a second after tine weeding. A final population count was also taken from the hand harvested area at site one. Effect of population thinning from tine weeding is summarized in table 1.
Table 1. Effects of population thinning from tine weeding
15” rows 30” rows
Site 1 8 % 9%
Site 2 11 % 18 %
Average of both sites 10% 15%
Populations were reduced 8% and 11% in the 15 inch rows and 9% and 18% in the 30” rows. Site 1 was a fairly level field while site two was a steeper hillside. In addition to the thinning from tine weeding there was some “cultivation blight” at both sites which sporadically destroyed several feet of row before the cultivator was positioned well between the rows.
See Tables 2a, 2b, 3a and 3b for a summary of population thinning by site, replication and treatments.
Table 2a. Percent Change in population after tine weeding (Site 1)
Replication Row Spacing % Change in population
1 15” 16
2 15” 6
3 15” 7
4 15” 5
Overall Average 8
Table 2b. Percent Change in population after tine weeding (Site 1)
Replication Row Spacing % Change in population
1 30” 11
2 30” 4
Overall Average 9
Table 3a. Percent Change in population after tine weeding (Site 2)
Replication Row Spacing % Change in population
1 15” 13
2 15” 12
3 15” 6
4 15” 12
Overall Average 11
Table 3b. Percent Change in population after tine weeding (Site 2)
Replication Row Spacing % Change in population
1 30” 13
2 30” 23
3 30” 19
4 30” 18
Overall Average 18
The cultivated plots had much less weed pressure and the plants appeared more vigorous through the season and displayed less stress at the end of the season. The plants in the narrow row plots were shorter, more spindly and had smaller ears.
Yield Summary
At site 1 the corn was hand harvested as dry shell corn. The site yielded an average of 85 bushels/acre in the 15” rows and 131 bushels/acre in the 30” rows. The differences in yield were statistically significant, validated by analysis of variance.
At site 2 the crop was harvested as silage. Since a single repetition was harvested the entire length of the field and then weighed we had only 4 repetitions per row spacing treatment compared to 32 repetitions with cover crop subplots under hand harvest. There was no statistically significant difference in yields between the two row spacings. Both treatments averaged 7 tons of silage at 35% drymatter.
Tables 4 – 9 summarize yield results by site.
Table 4. Dry Shell Corn Yield Site 1 Narrow Rows (15”)
Dry Shell Corn Yield Site 1 Narrow Rows (15”)
Replication Bu/ac @ 15.5% moisture
1 52.8 a
2 46.2 a
3 32.9 a
4 34.4 a
Table 5. Dry Shell Corn Yield Site 1 Wide Rows (30”)
Dry Shell Corn Yield Site 1 Wide Rows (30”)
Replication Bu/ac @ 15.5% moisture
1 129.1 b
2 117.9 b
3 113.3 b
4 106.9 b
Table 6. . Summary of Average Dry Shell Corn Yield by Treatment (Site 1)
Average Grain Yield by Treatment Bu/ac @ 15.5% moisture
15 “ rows 40.6 a
30 “ rows 123.2 b
Table 7. Silage Yield Site 1 Narrow Rows (15”)
Silage Yield Site 2 Narrow Rows (15”)
Replication T/ac @ 35% DM % DM at harvest
1 10.2 a .29
2 7.4 a .24
3 5.0 a .29
4 6.3 a .29
Table 8. Silage Yield Site 1 Wide Rows (30”)
Silage Yield Site 2 Wide Rows (30”)
Replication T/ac @ 35% DM % DM at harvest
1 5.6 a .27
2 7.4 a .29
3 8.8 a .27
4 6.4 a .24
Table 9. Summary of Average Silage Yield by Treatment
Average Silage Yield by Treatment T/ac @ 35% DM
15 “ rows 7.2 a
30 “ rows 7.0 a
Field Conditions and Challenges
We did not have a fixed 15” row planter available to us. To establish 15” rows we used a planter with 30” row spacings doubled back on itself. We had a very experienced operator plant both sites. Unfortunately the row spacings were very inconsistent varying from 0” to 15” with rows having an average spacing of 4-6”.
Our establishment of 15” rows was not precise enough for a completely fair evaluation in this study.
The two field sites had different soil types and different weed complexes. Early in the season site 1 had a fairly strong population of quackgrass. It seemed to fade later in the season and was replaced by ragweed as the dominant weed. Three tine weedings and two cultivations were done at this site. Cover crops were seeded about 1.5 weeks after the final cultivation, July 11-13. At that time the corn was chest high and the annual broadleaf weeds nearly as tall. The cover crops were able to establish in the cultivated plots but weed cover in the narrow rows prevented consistent germination. Where the cover crop seedlings were detectable in the field their growth remained insignificant. They were completely overwhelmed by the crop and weeds in the field.
Site two had even less success with cover crop establishment because the ground cover with weeds, mainly annual grasses and particularly crabgrass prevented the seeds from even making seed to soil contact. Cover crops were seeded July 14-15. The corn varied from knee to waist high. Rain and wet soil conditions prevented Site 2 from completion of a second row crop cultivation. The crabgrass had the chance to establish and the tine weedings could not adequately disturb the root systems. Weed competition was significant in the field and the corn looked stressed. The final yield was the same for both row spacings at 7 tons of silage/acre at 35% dry matter.
The study was performed and data collected. Anecdotal information drove the interest in a formal study. One of the cooperating farmers had tried narrow row corn on a hillside where cultivation was impractical and was satisfied with the yield results. That experience spurred interest in a closer look at the practice.
In this experiment it was very clear that there was significantly less yield where there was only tine weeding and no cultivation. One factor that likely impacted the poorer yield outcome of the trial versus the cooperating farmer's previous years positive experience was that 13.39 inches of rainfall was recorded at the farm from May through July of the trial year versus only 10.12 inches for the same period the prior year. Weed pressure was much greater in the trial year than during the previous year's narrow row crop. This experience indicates that abundant / excessive moisture tends to increase weed growth which impacts yields in a management system that does not use row crop cultivation or when weather prevents timely row crop cultivation.
Education & Outreach Activities and Participation Summary
Participation Summary:
A field day was held Sept. 2 in Truxton, NY which is located in south central NY. The field day was hosted by the Arnold family (Twin Oaks Dairy), one of the cooperating farms in this study. Bob LaFrancois, a provider and mentor of mechanical weed control was a featured speaker. Bob explained the differences among makes and models of tine harrow weeders, their best fit for crop and weed in the field and adjustments for best operation. His talk was enlightening, practical and very valuable. A pictorial display provided a history of the plots. We toured the SARE plot, the organically managed forage and grain fields of the Arnolds and a demonstration plot of grazing brassicas. There were 28 attendees. Many of the attendees traveled several hours to attend the field day. Kathie Arnold gave her time to talk with attendees one on one after the tours were through. Judging from the discussions, questions and feedback from attendees it was a valuable workshop with time well spent.
An excerpt of this final SARE report will be published in the Area Extension newsletter, the Dairy and Field Crops Digest. The article will be made available to the Northeast Organic Dairy Producer’s Association and NOFA-NY for publication in their newsletters.
Project Outcomes
Not applicable to this study.
Farmer Adoption
Based on the results of this experiment I would not recommend relying on narrow row spacing as an adequate means of weed suppression. I would caution anyone that wants to further evaluate narrow rows and weed suppression to try it on limited acreage under their field conditions. The results in any year will vary based on skill of the person operating the cultivation equipment, weed pressure, growing season and weather conditions. This study also suggests that an increase in seeding rate of at least 10% is needed to compensate for population thinning from weed control passes.
Areas needing additional study
Timely weed control is a critical factor for achieving optimal yields in corn production. From my experience and observations during this field study, cover crops would need to be established the season prior and winter killed or suppressed in some other way, such as rolling at the time of flower, to provide significant weed control. The attempt to establish cover crops mid-season as we did in this trial can only be described as too little too late. The window where weed control needs to be done had passed.
Rolling cover crops has shown success in areas further south of here, such as in Pennsylvania, where the growing season is longer. There have been some on-farm research and demonstrations in Central NY that have shown some positive results (Skaneateles Lake Watershed). Determining the best cover crop species, pairing with crop and window for suppressing the cover crop by rolling need further exploration. An alternative system for weed control needs to be explored because the cover crop doesn’t allow mechanical weed control.