By Christine Ziegler-Ulsh
The 2011 growing season was not a kind one for many farmers in southeastern Pennsylvania and other parts of the east coast, primarily due to rainfall patterns. Rodale Institute received significantly more rain than normal for the full year, particularly in the spring and late summer. However, June and July were extremely dry (see Figure 1 below). While those months appear to have received just-below-normal rainfall, that rain fell primarily before June 3 and after July 28, leaving a seven-week stretch in June and July – a prime plant development period – during which the region received essentially no rainfall.
That meteorological picture provided an interesting backdrop against which to compare organic no-till production systems with standard-tilled organic systems. For a fourth year in a row, Rodale Institute staff worked on the Integrated Organic Program (IOP) project titled “Developing Carbon-Positive Organic Systems through Reduced Tillage and Cover Crop-Intensive Crop Rotation Schemes” (lovingly referred to as the IOP Project).
This project began in 2008 in collaboration with Iowa State University (the lead partner), Michigan State University, University of Minnesota, North Dakota State University, and University of Wisconsin to test and compare organic tilled and no-till corn and soybean production rotations in diverse grain-producing regions of the U.S. The goal of the project is to identify organic grain production systems that best enhance soil quality by maximizing cover to minimize erosion, improve soil ecology and biological processes, reduce costs of production, and produce reliably high yields.
Early data from the first two years of the project were presented at Rodale Institute’s Annual Field Day in July 2010 and in the article Applied no-till for carbon-positive farming. Now, with another year of data, the project is coming to a close, and it’s time to summarize what we learned during the 2011 farming season. How those data compare to those of 2009 and 2010 paints a longer-term picture of the successes and challenges of organic no-till production.
Wet DRY Wet
The crazy weather of 2011 created challenges for both the tilled and no-till organic systems. We were growing corn in a hairy vetch cover crop and soybeans in rye. Our monthly soil moisture measurements illustrate the impact of the strange precipitation patterns on moisture availability for the crops (see Figure 2 below). Soil moisture in mid-July was roughly half of what it was in mid-June (just at the start of the dry spell) and mid-August (after the rains returned). What’s more, at all those points, the soil moisture content appeared to be unaffected by the presence or absence of a cover crop mat. That drop in moisture availability at critical June/July growth stages in both the corn and soybean lifecycles may almost certainly have had some impact on yields for both the crop systems.
The pattern of moisture availability in 2011 (wet spring, dry early summer, wet late summer) also impacted the growth and development of our hairy vetch cover crop. Figure 3 charts the amount of cover crop biomass grown and collected from the four different treatments at two different times: 1) at the beginning of the growing season in late March (the striped bars) and 2) just prior to the point when the cover crop was killed (the solid bars). You can see that in late March, the hairy vetch in the tilled and no-till corn plots (green bars) had higher biomass growth than the rye in the tilled and no-till soybean plots (yellow bars).
In the tilled corn and soybeans plots (the darker green and yellow bars), the cover crops had to be killed (in this case, plowed under) about a month earlier than in the no-till plots (the dates at the top of each bar denote the collection date). The plow can only successfully turn under and incorporate cover crops of a certain size. In this case, we must turn the rye and vetch cover crops under when they’re just about knee-high or else the biomass becomes more than the plow can handle. The rye cover crop reached appropriate plowing height on April 18, and the vetch reached it on May 12. (Vetch vines are not as thick and tough as rye stalks and grow in a more prostrate habit.) On those dates you can see that the rye doubled its biomass from the March cutting and the vetch almost tripled its biomass from the earlier date. These data are exactly what we expected and represent normal cover crop growth.
However, we were surprised by the cover crop biomass we collected at the later rolling dates in the no-till plots (the lighter solid green and yellow bars). In order to kill the cover crops with the roller/crimper, we have to wait about a month longer than the plow date, when the cover crop is entering into its seed-producing phase – about 50% flowering for the vetch and post-anthesis “soft dough” stage for the rye. If we try to roll before that point, the cover crop won’t die and will start to re-grow after the rolling. If we wait too late after that point, the cover crop will produce viable seed that will re-seed itself and start to grow as a weed. For organic no-till systems, proper timing of cover crop rolling is absolutely critical to success, but that timing forces the grower to delay planting past normal dates for tilled systems.
Nitrogen and carbon
As expected, the rye put on a massive amount of growth during the month of May, quadrupling its biomass from the plowing date a month earlier as it grew from knee-high to over 6 feet tall (a statistically significant difference, as referenced by the letters on those bars, Figure 3). But the hairy vetch did not! In fact, the vetch didn’t increase in biomass at all from mid-May to mid-June, a result we haven’t seen here at Rodale Institute in all our many years of cutting hairy vetch cover crop biomass.
In most years, hairy vetch produces just about as much biomass as rye does during the month of May. The added biomass is what allows hairy vetch to create a solid rolled cover crop mat to suppress weeds and add nitrogen to the soil. By not generating that extra biomass in May of 2011, the rolled vetch plots were less likely to create the weed-suppressing mat or supply the nitrogen that the ensuing corn crop would need to survive and thrive.
This concern is borne out in the data. Figure 4 shows the amount of nitrogen and carbon that the tilled or rolled cover crop biomass contained when it was killed. (The solid bars represent the total amount of biomass, the checkered bars represent nitrogen content, and the dotted bars represent carbon content.) As a grass, rye is usually very high in carbon and does not contain enough nitrogen to be considered a viable nutrient input. This holds true for the IOP trial in 2011, as the nitrogen input of the rye doubled over the month of May (from 45 to 80 lb per acre) while the carbon content almost septupled.
In many years, we also see nitrogen and carbon input in hairy vetch double over the month of May, but not in 2011. As you can see, the vetch biomass in this trial barely increased over the month, and though the carbon input of that biomass did show a small but significant increase, the nitrogen input actually decreased a little (though not significantly).
The solid bars represent the total amount of biomass, the checkered bars represent
nitrogen content, and the dotted bars represent carbon content.
We cannot definitively say why the vetch performed so poorly over the month of May, but our observations lead us to believe that the incessant wet weather that month (10 inches of rain compared to the average 4.2 inches) caused the thick, vigorous vetch to begin to rot at ground level a lot sooner than it normally does. Cutting mature vetch in late May is always a slimy spelunk of a job, but the amount of rot in the understory of the 2011 vetch was excessive. This premature plant tissue breakdown likely prevented the vetch crop from fully developing its usual nitrogen or carbon resources.
This weak vetch growth would lead one to conclude that the no-till corn plots (planted into the rolled vetch) might be the weediest of the four treatments. However, the same spring rains that caused the vetch to rot also prevented our field operators from planting the tilled corn as early as usual, and it also kept them from doing the normal and necessary early rotary hoeing and tine weeding for weed management.
Our tilled corn populations averaged 22,300 plants per acre, while our no-till corn averaged 32,500 plants per acre (our corn population goal is between 30,000 and 35,000; Figure 5). This difference was not significant, due to variations from plot to plot, but the impact was still clear when looking at the weed biomass and yields.
Perennials versus annuals
The total weed biomass was significantly higher in the tilled corn plots than in the no-till corn plots (and higher than the weed biomass in either soybean treatment; Figure 6), and this weed biomass skewed heavily toward annual weeds. Similarly, in the soybean plots, we found that the no-till system produced significantly more perennial broadleaf weeds than the tilled system.
These data reflect an outcome that we’ve seen consistently in comparing the tilled and no-till organic systems over the years. The tilled systems tend to favor growth of annual weeds (in our region, those weeds are primarily pigweed, ragweed, lambsquarters and foxtail), and the no-till systems favor the development of perennial weeds (here that includes Canada thistle, bindweed and dock). Regular tillage in the tilled systems continuously mines new annual seeds from the soil’s weed seed-bank and brings them to the surface to germinate but tends to break up the roots of perennial weeds before they can get too well established. Less frequent tillage makes a poorer soil surface for annual weed seed germination but provides the perfect environment for the solid establishment of deep-rooted, rhizome-reproducing perennials.
So after all this talk of rain, cover crops, soil and weeds, the most important question on the minds of all farmers and ag researchers is: What kinds of yields did all these factors conspire to create? The answer is, unsurprisingly, poor ones. Figure 6 summarizes the results quite succinctly. Our yield goals for corn and soybeans in this region are 150 and 40 bu/ac, respectively, and the only system that came close to reaching its yield goal was the tilled soybean system, which averaged 39 bu/ac over the four replications. Despite similar weed pressure, the no-till soybeans yielded only half of the tilled beans’ total. And, despite the fact that the no-till corn system had higher corn populations and lower weed pressure, it yielded less than one-third of its yield goal (about 40 bu/ac), while the tilled corn yielded about two-thirds of its yield goal (about 95 bu/ac).
We suspect that part of the poor performance of the no-till corn was lack of nitrogen input from the poor vetch growth in the month of May coupled with a loss of moisture at critical growth and grain development phases. (The no-till corn was planted 12 days later than the tilled corn, right at the beginning of the dry spell.)
However, possible reasons for the poor performance of the no-till soybeans are less clear. The two soybean treatments were planted within a few days of each other in late May, just a bit before the dry spell began, and weed pressure in both treatments was about the same in terms of total biomass (though the no-till beans had more perennial broadleaf weeds).
One possible answer to this no-till soybean conundrum may lie in the rye cover crop biomass, which was thick and plentiful enough to provide good weed suppression, but did not seem to improve the moisture level of the soil beneath it, and in fact may have removed some of the soil moisture for its own spring growth. While the thick rye biomass cover provided no moisture advantage, it forced the no-till beans to grow a few inches taller (getting a little more leggy) than their tilled counterparts just to break through the mat and put out their cotyledons. While this did not appear, visually, to affect the bean growth over the course of the season, this extra early “leggy-ness” may have cost the soybean plants some vigor and may also have caused some minor end-of-season lodging that would make it more difficult for the combine to pick up all the beans. (The rye biomass mat itself may have also interfered with the combine bean pick-up as well.)
Two other parameters that may have influenced our 2011 IOP results are soil nutrient and organic matter levels. Soil samples from our project field were collected in the fall of each field season (five years in all), and all the soil samples from each project site were sent to the National Soil Tilth Lab in Ames, IA, for analysis. Data from the 2011 samples haven’t been received from the lab as yet. There is a chance the data may further clarify issues and differences between the tilled and no-till systems. If so, we will be sure to update this article.
Researchers at the other IOP project sites in Iowa, Minnesota, Michigan, Wisconsin and North Dakota are summarizing their 2011 data at this time, and these data are being pooled by project leader Dr. Kathleen Delate from Iowa State University to summarize the overall results of this project. Based on the data from the 2008-2010 field seasons, coupled with observations from the first half of the 2011 season, the project team joined together again in November. We discussed what practices worked and what didn’t, taking into consideration location and timing. We’re now applying for a new funding stream to support research improving the practices begun during this project and to expand them to vegetable production systems in addition to the grain crop systems.
The total project results summary for both the Rodale site and for all the project sites as a whole, as well as the group’s plans for future organic no-till research, will be covered in our next article coming later this summer. Stay tuned!