Using Crop Diversity in No-till and Organic Systems to Reduce Inputs and Increase Profits and Sustainability in the Northern Plains
A major deterrent to adopting no-till or organic crop production in the Northern Great Plains is concern about weed management problems during the transition from conventional systems and moisture conservation associated with crops as alternatives to fallow. A large-scale experiment was established at Moore, MT to compare no-till and organic systems to conventional small grain production systems (Table 1). A similar systems experiment was established at Bozeman, MT (Table 2). The goal was to evaluate farm profitability and sustainability within a system that increases crop diversity and reduces off-farm inputs. Thus we have been quantifying production, profitability, and weed population dynamics among the production systems in the experiment at Moore and on farms in north central Montana.
1. Understand weed dynamics to predict species shifts within organic and no-till production systems.
2. Determine crop performance and water use efficiency within organic and no-till systems.
3. Quantify input levels and costs for organic and no-till systems, including both purchased and operator supplied inputs.
4. Quantify profitability (net return) for organic and no-till systems.
5. Educate producers on potential benefits of organic and no-till systems through dissemination of research results.
Andrew Hulting’s Ph.D. thesis research (originally advised by Bussan and now advised by Maxwell and Miller) associated with Objective 1 has continued (beginning in year 2000) with weed population data collection from the cropping system experiments at Moore, MT (Table 1) and Bozeman, MT (Table 2). Including field pea or canola (conventional or herbicide tolerant) in a crop rotation either consecutively or alternating with a small grain crop significantly reduces populations of the grass weed downy brome (Bromus tectorum) compared to those populations in small grain-fallow rotations. However, including broadleaf crops in rotations has led to population increases in some broadleaf weed species present at the site such as wild buckwheat (Polygonum convolvulus) and pinnate tansymustard (Descurainias sophia). Population densities of these two species have increased mainly in the reduced-input treatments of diversified rotations where soil-applied broadleaf herbicide products applied in the fall have been eliminated. We have identified and quantified some useful metrics that describe wild oat (Avena fatua) population growth or decline across the conventional, reduced, and organic treatments. Wild oat populations growth rate (l) under conventional herbicide management has been increasing annually, while the growth rate of populations under organic management have stabilized or are decreasing. The spatial extent of wild oat populations in all three management regimes (conventional, reduced input, and organic) has increased annually since 2000. Wild oat populations now occupy 8-10 % of the 3 m X 7.3 m area that is mapped annually in each plot. This percentage of occupied area is lower than that which was originally hypothesized for the 5th year of the study. In general, wild oat populations in all three management regimes remain spatially aggregated near the source populations, but the trend in the data suggests that all populations could be characterized as becoming more spatially random with greater variation (less predictable) in the conventional and reduced-input treatments. We continue to refine hypotheses related to the mechanisms driving this trend.
The second objective was assessed using data from the Bozeman, MT field experiment where Miller was the primary PI. The 2003 crop year completed the first cycle of the 4-yr rotations. All phases of all rotations are present each year. Plot size is 7.3 x 15.2 m. All plots were seeded with a Fabro no-till disc seeder with adjustable row spacing and fertilizer side-banding capability (separate adjustable gang of disc openers). All no-till plots were seeded at a 26-cm row spacing, except corn and sunflower, which were seeded at a 46-cm row spacing. All organic plots were seeded with a 13-cm row spacing. Pesticide application was made in no-till rotations with a shrouded 7.6-m wide sprayer to control pesticide drift. All plots were harvested with a Wintersteiger plot harvester Yield samples were obtained from a central 1.5-m wide harvest pass and then the rest of the plot was harvested in bulk. In no-till rotations, cereal crops were cut at 50% of plant height, while in organic systems they were cut at 25% of plant height. Two soil moisture samples were obtained from the center of each plot to a depth of 1.82 m and pooled in 30.4-cm increments down the profile for soil water calculations. All grain samples are processed to remove weeds, chaff, and other foreign matter prior to weighing for yield determination. The rotations are described in detail below, with yield and precipitation use efficiency tabulated by phase and year. All productivity parameters are on a dry matter basis in units of t ha-1. For economic analyses, representative grain moisture contents were assumed for all crops. The growing season and crop year rainfall summary appears in Table 1 below. 2000-03 ranged from 49 to 92 mm drier than the long-term average of 380 mm for this location. Growing season rainfall patterns in all four years were consistent with long-term patterns, with seasonal peaks occurring in May and June.
The precipitation pattern at Bozeman is characterized by ample winter snowfall and early spring rain which effectively recharges soil water to a depth greater than is common to most cropland in Montana. The deep rooting pattern of sunflower has been well suited to this moisture pattern, making impressive yields in the face of severe midsummer drought. In contrast, the shallow rooting of pea has been ineffective. However, the deep soil water use by sunflower creates a hangover effect in the rotation. Spring pea yield following sunflower yielded 25% less than that following spring wheat in 2001-03. Worse, it appears that the hangover lasts two years following sunflower because winter wheat yields on pea stubble were 17% less when sunflower was in the rotation two years earlier instead of winter wheat. Wheat yields following pea averaged 8% and 23% greater than that following canola for spring wheat and winter wheat, respectively, consistent with earlier reports by Miller et al. (2002 and 2003b).
Soil water use patterns are widely held to be important in semiarid cropping systems. Where stored soil water is available, deep rooted crops may continue to make yield in the face of drought, but may cause negative yield effects on subsequent crops. Conversely, shallow rooted crops may not use deep stored soil water during drought, but conserve water for subsequent crops. Crops differed strongly in soil water extraction for each year where it was measured (2001-2003). Crops ranked consistently in the following manner: Pea < Lentil << Spring wheat < Winter wheat << Sunflower (see Figure 1), consistent with earlier reports by Miller et al. (2002 and 2003a). There was no difference in water use between winter wheat grown under no-till or grown organically. Canola water use was variable from year to year, perhaps indicating a plastic rooting response for this crop. Soil water extraction by corn was similar to wheat but proso millet was greater than spring wheat in 2001. Proso millet was initially included in the No-Till Diverse rotation to help balance soil water use after deep-rooted sunflower, but when it became apparent that millet was not a shallow water user, it was replaced with corn. Observations from related crop studies in 2000 confirmed soil water use by proso millet to be equal to or greater than spring wheat.
There are two parallel efforts to evaluate the economic returns to organic and no-till crop rotations. The first method evaluates controlled plot studies on university carried out by researchers involved in this project. The second utilizes surveys of no-till and organic producers in central Montana to ascertain their rotational experience concurrently with the plot studies.
Method 1. Data from Miller’s experimental plots on the Post Farm in Bozeman, MT have been evaluated for the period 2000-2003. Detailed data for input use, tillage practices, yields, and crop quality have been used to assess the relative net returns of three no-till rotations, one organic rotation, and a continuous crop-conventionally tilled wheat rotation. Returns for the organic rotation were evaluated under both representative organic premiums and under no premiums. The no-till rotations included a spring wheat-based rotation, a winter wheat-based rotation, and a diverse rotation.
The average per acre net returns (abstracting from returns to labor, land, management, and machinery ownership) from all five rotations considered for the Post Farm site are illustrated in Figure 1. Returns from the organic rotation with premiums are consistently higher than those for the competing rotations, largely due to the reduced herbicide and fertilizer inputs. The importance of the organic premiums is clearly shown by the differences in the net returns for the premiums case vs. the no premiums case in the organic protestations. The spring wheat-based and the diverse no-till rotations provide the highest average returns of the remaining rotations.
Additional data from experimental plots in Moore, MT is expected to be available early in 2004. The net returns from various crop rotations using data will be evaluated in a manner comparable to the analysis used for the Post Farm data.
Method 2. The plot data discussed above provide some very useful comparisons of returns from various rotations in a scientifically controlled setting. With these relatively new rotations, however, come numerous questions regarding how effectively an experimental study matches the practices and experiences of commercial producers. To address these questions, a parallel effort gathered survey data from a number of long-time organic and no-till producers in Central Montana. These producers’ yields and production practices were elicited for years 1999-2003. Of particular interest for these survey interviews were the responses by these producers to the extensive and severe drought conditions experienced during crop years 2001, 2002, and, to some extent, 2003. These interviews were completed in December 2003, with some follow-ups to be carried out early in 2004.
The crop input and yield data will be evaluated in 2004. The qualitative data on responses to drought conditions revealed that both no-till and organic producers generally reduced their cropped acreage (utilized more fallow). No-till producers experienced less wind erosion than did organic producers and neighboring farms using conventional tillage. This wind erosion varied considerably across the region, with some areas experiencing very little erosion and others experiencing considerable erosion.
Impacts and Contributions/Outcomes
The predominant small grain cropping system of the Northern Great Plains has utilized a crop fallow rotation under an assumption of moisture conservation. More sustainable systems may include the use of more diversified and organic systems that reduce inputs. The transition to more sustainable systems represents a challenging step for producers and requires prediction of how crops and weeds will perform. Weed population dynamics in response to more sustainable agricultural systems on the Northern Great Plains is critical to implementation of these systems. This study has already begun to show significant differences in weed behavior under different crop management and promises to allow prediction of weed responses to a wide range of management approaches from organic to high-input conventional. To our knowledge, there are no other studies that have simultaneously measured spatial and temporal dynamics of weed populations. Thus, there has been limited ability to predict the economic thresholds or distributions that determine optimum weed management under most conditions.