The adoption of conservation-tillage practices and glyphosate-resistant (GR) crops over the last several decades has led to an increased reliance on herbicides as one of the primary methods of weed control (Culpepper et al. Reference Culpepper, York, Batts and Jennings2000; DeVore et al. Reference DeVore, Norsworthy and Brye2013; Krausz et al. Reference Krausz, Kapusta and Matthews1993; Young Reference Young2006). Glyphosate-resistant crops have been rapidly adopted since their release in 1996 and have enabled producers to simplify weed management by providing control of a broad spectrum of common weeds with little or no injury to the crop (Fernandez-Cornejo and Mcbride Reference Fernandez-Cornejo and Mcbride2002). In 1997, only 17% of US soybean hectares were planted with herbicide-resistant varieties (Fernandex-Cornejo and Wechsler Reference Fernandex-Cornejo and Wechsler2015). By 2015, 94% of soybean hectares were planted with herbicide-resistant varieties, with the vast majority of those being GR (Fernandex-Cornejo and Wechsler Reference Fernandex-Cornejo and Wechsler2015). The continuous use of glyphosate on millions of hectares has led to the selection of GR weed biotypes around the world (Heap Reference Heap2016). The United States currently has 14 GR weed species, including 3 Amaranthus species: Palmer amaranth (Amaranthus palmeri S. Wats.), spiny amaranth (Amaranthus spinosus L.) and waterhemp (Amaranthus rudis Sauer). Among the 14 GR weeds, waterhemp and Palmer amaranth are consistently ranked as two of the most common and persistent weeds in southern and midwestern United States crops, especially in soybean (Beckie Reference Beckie2006; Bradley Reference Bradley2013; Heap Reference Heap2016; Legleiter and Johnson Reference Legleiter and Johnson2013; Schultz et al. Reference Schultz, Weber, Allen and Bradley2015b; Webster and Nichols Reference Webster and Nichols2012). The evolution of glyphosate resistance in weeds like Palmer amaranth and waterhemp has increased production costs in soybean and complicated weed management dramatically (Legleiter et al. Reference Legleiter, Bradley and Massey2009; Mueller et al. Reference Mueller, Mitchell, Young and Culpepper2005).
Soybean yield losses of 43% have been reported after 10 wk of interference by waterhemp at densities of 89 to 362 plants m−2 (Hager et al. Reference Hager, Wax, Stoller and Bollero2002). Palmer amaranth at a density of 8 plants per meter of row that emerged with soybean reduced yields by 79% (Bensch et al. Reference Bensch, Horak and Peterson2003). Currently, Palmer amaranth populations with resistance to six different herbicide modes of action have been confirmed: 5-enolpyruvyl-shikimate-3-phosphate (EPSP) synthase–inhibitors, microtubule-inhibitors, photosystem II–inhibitors, acetolactate synthase (ALS)-inhibitors, protoporphyrinogen oxidase (PPO)-inhibitors, and 4-hydroxyphenylpyruvate dioxygenase (HPPD)-inhibitors (Heap Reference Heap2016). In addition, waterhemp with resistance to six different herbicide modes of action has been reported in the United States: synthetic auxins and EPSP-, ALS-, photosystem II–, PPO-, and HPPD-inhibitors (Heap Reference Heap2016).
The increase in the occurrence of multiple-herbicide resistance in Amaranthus species and other weeds illustrates the need for producers to diversify their weed management practices (Bradley Reference Bradley2013; Norsworthy et al. Reference Norsworthy, Ward, Shaw, Llewellyn, Nichols, Webster, Bradley, Frisvold, Powles, Burgos, Witt and Barrett2012). Cultural control practices such as tillage can significantly impact weed populations (Norsworthy et al. Reference Norsworthy, Ward, Shaw, Llewellyn, Nichols, Webster, Bradley, Frisvold, Powles, Burgos, Witt and Barrett2012). Tillage has a large impact on the vertical distribution of weed seed in the soil profile and on weed emergence (Cousens and Moss Reference Cousens and Moss1990; Roberts Reference Roberts1963; Starica et al. Reference Starica, Burford, Allmaras and Nelson1990). Tillage implements that provide deep inversion of the soil place weed seed low enough in the soil profile to prevent successful germination and emergence (DeVore et al. Reference DeVore, Norsworthy and Brye2013). Shaw et al. (Reference Shaw, Culpepper, Owen, Price and Wilson2012) defined inversion tillage as tillage that flips over a layer of soil (often 15 to 30 cm), burying surface residues in the process. Moldboard plowing, a type of inversion tillage, not only kills plants but also can bury >95% of weed seeds at a depth from which most cannot emerge (Douglas and Peltzer Reference Douglas and Peltzer2004; Morris et al. Reference Morris, Miller, Orson and Froud-Williams2010). Swanton et al. (Reference Swanton, Shrestha, Knezevic, Roy and Ball-Coelho2000) found that moldboard plowing resulted in 63% of weed seeds being concentrated at a depth of 10 to 15 cm. In contrast, Clements et al. (Reference Clements, Benoit, Murphy and Swanton1996) and Pareja et al. (Reference Pareja, Staniforth and Pareja1985) both found that no-tillage systems result in the largest concentration of weed seed being contained in the uppermost layers of soil. However, soil type has also been shown to affect the vertical distribution of weed seed caused by tillage (Swanton et al. Reference Swanton, Shrestha, Knezevic, Roy and Ball-Coelho2000). Diversified integrated weed management strategies, including the incorporation of cultural practices such as tillage, narrow row spacing, and increased seeding densities, are among the best management practices currently recommended for the prevention, mitigation, and management of herbicide-resistant weed species (Beckie Reference Beckie2006; Norsworthy et al. Reference Norsworthy, Ward, Shaw, Llewellyn, Nichols, Webster, Bradley, Frisvold, Powles, Burgos, Witt and Barrett2012; Schultz et al. Reference Schultz, Myers and Bradley2015a).
The initial effectiveness and simplicity of the GR cropping system led many producers to rely solely on POST herbicide applications for weed control in soybean (Powles Reference Powles2008; Young Reference Young2006). As a result, Palmer amaranth and waterhemp have evolved resistance to herbicides that act at many different sites of action, but, at the current time, neither of these species has evolved resistance to glufosinate (Heap Reference Heap2016). If used appropriately, glufosinate remains an effective POST option for the control of Amaranthus species in glufosinate-resistant soybean (Heap Reference Heap2016; Norsworthy et al. Reference Norsworthy, Griffith, Scott, Smith and Oliver2008). However, it has been shown that repeated POST-only herbicide applications can lead to the selection of herbicide-resistant weed biotypes (Bradley Reference Bradley2013; Powles Reference Powles2008). The use of residual herbicide applications, PRE and POST, has been shown to reduce weed densities while also reducing the likelihood of herbicide resistance (Beckie Reference Beckie2006; Bradley Reference Bradley2013; Legleiter et al. Reference Legleiter, Bradley and Massey2009; Schultz et al. Reference Schultz, Myers and Bradley2015a).
The effect of tillage systems and residual herbicide programs on the control of GR Amaranthus species in glufosinate-resistant soybean has not been researched extensively over a broad range of geographies and soil types. The objectives of this research were to 1) determine the effect of four tillage systems (deep tillage, conventional tillage, minimum tillage, and no tillage), with and without a residual herbicide program, on season-long emergence of Amaranthus species in glufosinate-resistant soybean, and to 2) determine the effect of these four tillage systems on the vertical distribution of Amaranthus seed in the soil profile.
Materials and Methods
A field study was conducted in 2014 and 2015 at sites in Randolph County, Missouri; Boone County, Missouri; Washington County, Arkansas; St. Claire County, Illinois; Tippecanoe County, Indiana; Clark County, Ohio; Madison County, Tennessee; and Columbia County, Wisconsin (Table 1). Glufosinate-resistant soybean varieties of an appropriate maturity group were seeded at 321,000 to 432,000 seeds ha−1 in rows spaced 76 to 91 cm apart, depending upon location. Specific site information, such as previous crop production history, crop information, and soil type, is provided in Table 1. Monthly rainfall totals are presented in Table 2.
Table 1 Site characteristics for field trials conducted in 2014 and 2015.Footnote a
a Abbreviations: CEC, cation exchange capacity (meq per 100 g soil); OM, organic matter.
b Due to excessive soil moisture at and following harvest, some fall-tillage operations were not able to be completed until the following spring.
c Fayetteville, Arkansas. Arkansas Agriculture Research and Extension Center, University of Arkansas (36.092996°N, 94.173423°W). Site has been in small-plot research with conventional tillage for the past 50 yr.
d Belleville, Illinois. Belleville Research Center, Southern Illinois University (38.521476°N, 89.845294°W). Small-plot research for more than 7 yr. No deep tillage for at least 30 yr.
e Lafayette, Indiana. Throckmorton Purdue Agricultural Center, Purdue University (40.271114°N, 86.881163°W). Research plots with conventional tillage for at least 9 yr.
f Columbia, Missouri. Bradford Research and Extension Center, University of Missouri (38.898432°N, 92.216371°W). Small-plot research for 50 yr. Conservation agriculture at least 12 yr.
g Moberly, Missouri. Resistant Waterhemp Research Site, University of Missouri (39.302782°N, 92.369678°W). Continuous soybean production for at least 12 yr. Conservation agriculture the past 15 years. Confirmed presence of waterhemp populations resistant to glyphosate and protoporphyrinogen oxidase– and acetolactate synthase–inhibiting herbicides.
h South Charleston, Ohio. Western Agricultural Research Station, Ohio State University (39.8593°N, 83.66971°W). Small-plot research more than 20 yr. Conventional tillage annually.
i Jackson, Tennessee. West Tennessee AgResearch and Education Center, University of Tennessee (35.624655°N, 88.845096°W). No deep tillage for at least 30 yr. Mostly no-till practices.
j Arlington, Wisconsin. Arlington Agricultural Research Station, University of Wisconsin (43.307943°N, 89.350072°W). Agronomy research for more than 10 yr. Chisel plowing (25 cm) occurred once, two years prior to initiation of this study.
Table 2 Monthly rainfall (mm) from April through October in 2014 and 2015 at all trial locations. The 30-yr monthly rainfall averages are provided for comparison.Footnote a
a Abbreviations: avg, average.
b Fayetteville, Arkansas. Arkansas Agriculture Research and Extension Center, University of Arkansas (36.092996°N, 94.173423°W). Site has been in small-plot research with conventional tillage for the past 50 yr.
c Belleville, Illinois. Belleville Research Center, Southern Illinois University (38.521476°N, 89.845294°W). Small-plot research for more than 7 yr. No deep tillage for at least 30 yr.
d Lafayette, Indiana. Throckmorton Purdue Agricultural Center, Purdue University (40.271114°N, 86.881163°W). Research plots with conventional tillage for at least 9 yr.
e Columbia, Missouri. Bradford Research and Extension Center, University of Missouri (38.898432°N, 92.216371°W). Small-plot research for 50 yr. Conservation agriculture at least 12 yr.
f Moberly, Missouri. Resistant Waterhemp Research Site, University of Missouri (39.302782°N, 92.369678°W). Continuous soybean production for at least 12 yr. Conservation agriculture the past 15 years. Confirmed presence of waterhemp populations resistant to glyphosate and protoporphyrinogen oxidase–and acetolactate synthase–inhibiting herbicides.
g South Charleston, Ohio. Western Agricultural Research Station, Ohio State University (39.8593°N, 83.66971°W). Small-plot research more than 20 yr. Conventional tillage annually.
h Jackson, Tennessee. West Tennessee AgResearch and Education Center, University of Tennessee (35.624655°N, 88.845096°W). No deep tillage for at least 30 yr. Mostly no-till practices.
i Arlington, Wisconsin. Arlington Agricultural Research Station, University of Wisconsin (43.307943°N, 89.350072°W). Agronomy research for more than 10 yr. Chisel plowing (25 cm) occurred once, two years prior to initiation of this study.
j Thirty-year averages (1982 to 2011) obtained from the National Climatic Data Center (2016).
Treatments were arranged as a split-plot design with four replications, where tillage system was the main plot and herbicide program was the subplot, arranged in a randomized complete block design. Four tillage regimes were evaluated: 1) a fall tillage pass of a moldboard plow followed by a pass with a field cultivator in the spring, referred to as the deep-tillage treatment; 2) a fall pass with a chisel plow followed by a pass with a field cultivator in the spring, referred to as the conventional-tillage treatment; 3) a single pass of a vertical tillage tool in the spring, referred to as the minimum-tillage treatment; and 4) a no-tillage treatment that received a burndown herbicide treatment of paraquat (0.84 kg ha−1) near the time of the spring tillage treatments. Dates of major field operations for each site are provided in Table 1. Each tillage treatment received one of two herbicide programs: 1) a PRE application of flumioxazin (0.09 kg ha−1) followed by a POST application of glufosinate (0.59 kg ha−1) plus S-metolachlor (1.39 kg ha−1), referred to as the residual herbicide program, or 2) POST-only applications of glufosinate (0.59 kg ha−1), referred to as the POST-only herbicide program. The specific herbicide formulations utilized are listed in Table 3. All herbicide treatments were applied with a CO2-pressurized backpack sprayer calibrated to deliver 140 L ha−1. Treatments were applied at a constant speed of 5 km hr−1. PRE treatments were applied at or just prior to planting. POST application for the residual herbicide program was applied 21 d after planting, and the first POST application for the POST-only herbicide program was applied approximately 14 d after planting.
Table 3 Sources of materials used in the experiments.
a Abbreviations: EC, emulsifiable concentrate; L, liquid; SL, soluble (liquid) concentrate; WDG, water-dispersible granule.
Field Densities
Amaranthus species emergence was monitored every 14 d from planting up to the R6 soybean stage or senescence by counting all plants within two 1-m2 quadrats between the center two rows of soybean. The location of each quadrat was permanently marked to ensure counts occurred in the same area for the duration of the experiment. Immediately after each weed count was completed, the entire area was treated with glufosinate (0.59 kg ha−1) and then monitored for surviving weeds.
Vertical Seed Distribution
Vertical distribution of Amaranthus seed in the soil profile was determined by taking six 2.5 by 25 cm soil cores randomly from each plot utilizing 2.9-cm diameter soil recovery probes (AMS Inc., American Falls, ID). Soil probes were fitted with 2.5 by 30.5 cm acetate sleeves (AMS Inc.). Soil cores were taken after the spring tillage operations prior to planting and herbicide application. Soil cores from all sites were packaged with dry ice and shipped overnight to the University of Missouri–Columbia, where they were stored at −9 C until processing.
Each soil core was divided into following segments by soil profile depth, with zero representing the soil surface: 0 to 1 cm, 1 to 5 cm, 5 to 10 cm, 10 to 15 cm, 15 to 20 cm, and 20 to 25 cm. Each segment was placed as a topsoil layer in an individual 8 by 6 by 6 cm insert cell in a 28 by 56 cm greenhouse flat (Hummert International™, Earth City, MO) that had previously been three-quarters filled with commercial potting medium (Premier Tech Horticulture, Quakertown, PA). Plants were maintained in a greenhouse at 25 to 30 C, watered and fertilized as needed, and provided with artificial light from metal halide lamps (600 µmol photon m−2 s−1) simulating a 16-h photoperiod. Seedling emergence was monitored over a 3-mo period. Emerged weed seedlings were counted and identified to species every 14 d, then removed from the flats after counting. After 3 mo of monitoring in the greenhouse, the flats were removed and stored in the dark at −9 C for 3 mo of cold stratification. After cold treatment, flats were returned to the greenhouse and the soil in each cell was stirred by hand. Flats were monitored for weed emergence for an additional 2 mo using the same photoperiod and temperature as before.
To ensure that the density of Amaranthus was adequate to compare treatments, field count data were only included in the statistical analysis if cumulative Amaranthus densities were greater than 300 plants per trial (12 of 16 site-years), and vertical distribution data were only included in the statistical analysis if Amaranthus densities were greater than 32 plants per trial (5 of 14 site-years; no cores were taken from the Ohio location). Amaranthus species density data from the field and vertical distribution data from the soil cores were analyzed separately using the PROC GLIMMIX procedure in SAS® version 9.4 (SAS Institute Inc., Cary, NC). Count data were transformed using a negative binomial function to satisfy Pearson’s chi-square. Data were back-transformed for presentation. Replicate, tillage system, and herbicide program were considered fixed effects for the field count data. Replicate and tillage system were considered fixed effects for the soil core data. Site-year combinations were analyzed as if they were random samples taken from the same environment, and site-years with adequate weed densities were combined for analysis (Blouin et al. Reference Blouin, Webster and Bond2011; Carmer et al. Reference Carmer, Nyquist and Walker1989). Individual treatment differences were separated using Fisher’s protected LSD at P≤0.05. Significant differences were present in the field between tillage systems (P<0.0001; Figure 1), herbicide programs (P<0.0001), and tillage system – herbicide program combinations (P=0.04; Figure 2). Significant differences were also present in the greenhouse studies between tillage system and soil profile depth (P<0.0001; Figure 3). In Indiana, only two tillage systems were evaluated: conventional tillage and no tillage. Data from Indiana were analyzed separately using the same procedure that was used for the other sites. Significant differences were noted in the field studies in Indiana between tillage systems (P<0.0068) and herbicide programs (P<0.0001), but not between tillage system – herbicide program combinations (P=0.7133; data not shown). No significant differences were found between tillage system and soil profile depth for the soil cores taken from Indiana (P=0.2061; data not shown).
Figure 1 Influence of tillage method on Amaranthus species emergence across 10 site-years in Arkansas, Illinois, Missouri, Ohio, and Tennessee. Bars with the same letter are not different, LSD (0.05).
Figure 2 Influence of tillage treatment and herbicide program on Amaranthus species emergence in the field across 10 site-years in Arkansas, Illinois, Missouri, Ohio, and Tennessee. Min. till: minimum tillage; Conv. till: conventional tillage. Bars followed by the same letter are not different, LSD (0.05).
Figure 3 Influence of tillage method on the vertical distribution of Amaranthus species seeds in the soil profile. Results combined across the Missouri and Illinois sites. Each bar proportionally represents the average number of Amaranthus species that emerged from each soil core segment. Percentages followed by the same letter are not different, LSD (0.05).
Results and Discussion
The deep-tillage system resulted in a 62%, 67%, and 73% reduction in Amaranthus emergence compared with the conventional-, minimum-, and no-tillage systems, respectively, when averaged over all sites except Indiana (Figure 1). The conventional-tillage system resulted in a 28% reduction in Amaranthus species emergence compared with the no-tillage system, but emergence was similar in the minimum- and no-tillage systems. Across both years in the experiment conducted in Indiana, there was a significant difference between the conventional- and no-tillage systems. In Indiana the conventional-tillage system resulted in a 64% reduction in emergence compared with the no-tillage system when averaged over all years, with 22 and 60 emerged plants m−2, respectively. The reason for the difference in results between Indiana and the other sites with regard to the relative effectiveness of the conventional and no-tillage systems is not readily apparent. Previous research utilizing similar treatments found that tillage systems that utilize a chisel plow (conventional treatment in this study) resulted in the majority of weed seed remaining high in the soil profile, similar to no tillage (Ball Reference Ball1992; Clements et al. Reference Clements, Benoit, Murphy and Swanton1996; Pareja et al. Reference Pareja, Staniforth and Pareja1985; Yenish et al. Reference Yenish, Doll and Buhler1992). A reduction in Amaranthus species emergence as a result of deep tillage was also observed in a study in Arkansas comparing effects of deep-tillage and no-tillage treatments on Palmer amaranth emergence (DeVore et al. Reference DeVore, Norsworthy and Brye2013). DeVore et al. (Reference DeVore, Norsworthy and Brye2013) found that deep tillage in an early soybean production system reduced Palmer amaranth emergence by 97% compared with no tillage, while deep tillage in a full-season soybean production system reduced Palmer amaranth emergence by 70% compared with no tillage. The level of Palmer amaranth reduction reported by DeVore et al. is similar to the 73% reduction in Amaranthus species emergence observed in this multi-state study (Figure 1). In another study with Palmer amaranth in Arkansas, Bell et al. (Reference Bell, Norsworthy and Scott2015) found that 14 d after planting soybean in a deep-tillage treatment, Palmer amaranth densities were reduced by 94% to 95% in one year and 73% to 87% in another. Similar results have also been observed in Iowa, where moldboard and chisel plow treatments each decreased waterhemp emergence 4-fold compared with the no-tillage treatments (Leon and Owen Reference Leon and Owen2006). The reduction in Amaranthus species emergence in deep- and conventional-tillage systems compared with no-tillage systems can be explained by the less favorable conditions for germination and seedling establishment for small-seeded weeds like Amaranthus when the seeds are buried deeper in the soil profile (Felix and Owen Reference Felix and Owen1999; Hoffman et al. Reference Hoffman, Owen and Buhler1998; Webster et al. Reference Webster, Cardina and Norquay1998; Yenish et al. Reference Yenish, Doll and Buhler1992).
The residual herbicide program resulted in an 87% reduction in Amaranthus species emergence compared with the POST-only program, when averaged over all except the Indiana sites, with 13 and 103 plants m−2, respectively. For the Indiana sites, the residual program resulted in a 98% reduction in emergence compared with the POST-only program, with 5 and 232 plants m−2 emerged, respectively. These results are similar to those reported by Schultz et al. (Reference Schultz, Myers and Bradley2015a), where a PRE fb POST with residual program had greater waterhemp density reduction (99%) than two-pass POST-only applications of glufosinate (72%) when compared to the non-treated control. Similar results have been observed across 4 site-years in Missouri in a study of herbicide programs in glufosinate-resistant soybean, where a residual program resulted in 93% control of waterhemp while two-pass POST-only program using glufosinate resulted in only 74% control (Craigmyle et al. Reference Craigmyle, Ellis and Bradley2013). In a study with GR waterhemp in Missouri, Legleiter et al. (Reference Legleiter, Bradley and Massey2009) observed 97% and 98% GR waterhemp density reductions with residual herbicide programs, but less than 40% reduction with POST-only programs. Use of PRE herbicide applications of metolachlor plus metribuzin have also been shown to provide much higher economic returns than POST-only applications of glyphosate when GR weeds are present (Legleiter et al. Reference Legleiter, Bradley and Massey2009). These results support the recommendation to plant into fields that are weed-free and to keep them weed-free by utilizing residual herbicide applications before and/or after planting (Norsworthy et al. Reference Norsworthy, Ward, Shaw, Llewellyn, Nichols, Webster, Bradley, Frisvold, Powles, Burgos, Witt and Barrett2012), thus reducing the selection pressure for resistance to POST-only herbicides (Neve et al. Reference Neve, Diggle, Smith and Powles2003; Neve et al. Reference Neve, Norsworthy, Smith and Zelaya2011).
We observed an interaction between tillage system and herbicide program at all sites except the Indiana sites. The lowest emergence of Amaranthus species occurred with the combination of deep tillage and residual herbicide program (Figure 2). The combination of any tillage system with residual herbicides resulted in less Amaranthus emergence than the combination of any tillage system and the POST-only herbicide program. There were no differences in Amaranthus species emergence among the conventional-, minimum-, and no-tillage systems where the POST-only herbicide program was used. The deep-tillage system in combination with the residual herbicide program resulted in a 97% reduction in Amaranthus species emergence compared with the minimum-tillage system combined with the POST-only program, which resulted in the highest Amaranthus species emergence. Similar results were observed in an Arkansas study with an application of flumioxazin plus pyroxasulfone (3-[[5-(difluoromethoxy)-1-methyl-3-(trifluoromethyl)pyrazol-4-yl]methylsulfonyl]-5,5-dimethyl-4H-1,2-oxazole) applied PRE in combination with deep tillage (Bell et al. Reference Bell, Norsworthy and Scott2015). This combination resulted in >98% Palmer amaranth control in both years of the study, and the authors concluded that deep tillage resulted in fewer Palmer amaranth plants present at the times of the POST herbicide applications (Bell et al. Reference Bell, Norsworthy and Scott2015). DeVore et al. (Reference DeVore, Norsworthy and Brye2013) concluded that adding a residual herbicide program to a deep-tillage early soybean production system would further reduce Palmer amaranth emergence, providing even greater control. The results of this research support this conclusion, as the deep-tillage system with the addition of a residual herbicide program provided the greatest reduction of Amaranthus species emergence across the major soybean-producing area in the United States (Figure 2).
Vertical Distribution of Amaranthus Seed
The conventional-, minimum-, and no-tillage systems resulted in similar Amaranthus species emergence from all soil depths (Figure 3). These results were also observed in a separate analysis across both years in Indiana in a comparison of the conventional and no-tillage systems (P=0.2061, data not shown). The deep-tillage system resulted in lower Amaranthus species emergence at the 0 to 1 cm, 5 to 10 cm, 10 to 15 cm, and 15 to 20 cm depths compared with the other tillage systems. Seventy-two percent of the Amaranthus seeds in the deep-tillage system were located deeper than 5 cm. These results are similar to those reported by Nichols et al. (Reference Nichols, Verhulst, Cox and Govaerts2015), who utilized previous tillage research (Dorado et al. Reference Dorado, Del Monte and Lopez-Fando1999; Mohler Reference Mohler1993) to predict that moldboard plowing would place the majority of seeds within the 5 to 15 cm soil profile range. Swanton et al. (Reference Swanton, Shrestha, Knezevic, Roy and Ball-Coelho2000) also reported that moldboard plowing resulted in less uniform vertical weed seed distribution in locations with sandy soils, but that 63% of the seeds were concentrated at depths of 10 to 15 cm. The results from this study indicate that 28% of the Amaranthus seeds were placed in the top 5 cm of the soil profile by deep tillage. In contrast, 79%, 81%, and 77% of the Amaranthus seed in the conventional-, minimum-, and no-tillage systems was located in the top 5 cm of the soil profile, respectively. These results are comparable to other studies, conducted across multiple soil types, that found that the majority of weed seed was located in the top 5 cm of the soil after conventional-, minimum-, and no-tillage treatments (Clements et al. Reference Clements, Benoit, Murphy and Swanton1996; Pareja et al. Reference Pareja, Staniforth and Pareja1985; Swanton et al. Reference Swanton, Shrestha, Knezevic, Roy and Ball-Coelho2000; Yenish et al. Reference Yenish, Doll and Buhler1992). For example, Swanton et al. (Reference Swanton, Shrestha, Knezevic, Roy and Ball-Coelho2000) reported that 90% of the weed seedbank was concentrated in the top 5 cm of soil in a no-tillage system.
Based on the results of this research, deep-tillage systems utilizing a one-time inversion tillage implement, such as a moldboard plow, provide the greatest reduction in Amaranthus species emergence when compared with conventional-, minimum-, and no-tillage systems. However, deep-tillage systems are prone to increased soil erosion as well and can incur higher fuel and labor costs (Logan et al. Reference Logan, Baker, Davidson and Overcash1987). Reducing tillage intensity through conservation-tillage practices has been shown to reduce soil erosion and water runoff (Baumhardt and Lascano Reference Baumhardt and Lascano1996; Reeves Reference Reeves1997) and increase soil organic matter, soil water-holding capacity, the quantity and diversity of soil organisms, and water infiltration (Bruce et al. Reference Bruce, Langdale, West and Miller1992; Heisler Reference Heisler1998; Kemper and Derpsch Reference Kemper and Derpsch1981; Reeves Reference Reeves1997; Truman et al. Reference Truman, Reeves, Shaw, Motta, Burmester, Raper and Schwab2003). Therefore, producers will need to weigh all of the potential risks, costs, and benefits before making a decision on either system. The results from this research also illustrate the effectiveness of a residual herbicide program when used in either tillage system. The incorporation of a residual herbicide, particularly a PRE herbicide prior to soybean planting, has been shown to provide better control and density reduction of Amaranthus species, as well as greater economic benefit, than POST-only programs (Legleiter et al. Reference Legleiter, Bradley and Massey2009; Schultz et al. Reference Schultz, Myers and Bradley2015a). The combination of a residual herbicide program and a deep-tillage system provided the greatest reduction in Amaranthus species emergence throughout the season. These results support the recommendation to combine effective cultural practices with residual herbicide programs; using both together has been shown to be more effective than utilizing any one management technique alone (Norsworthy et al. Reference Norsworthy, Ward, Shaw, Llewellyn, Nichols, Webster, Bradley, Frisvold, Powles, Burgos, Witt and Barrett2012; Schultz et al. Reference Schultz, Myers and Bradley2015a). As herbicide-resistant Amaranthus species become more prevalent throughout US soybean production systems, integrating cultural practices such as tillage, where appropriate, with residual herbicide programs that utilize multiple, effective sites of action can provide substantial reductions in herbicide-resistant Amaranthus species in soybean.
Acknowledgements
We greatly appreciate the funding provided by the United Soybean Board. Additionally, the corresponding author would like to thank all of the many individuals from the cooperating institutions for their hard work and dedication to this project.
