Introduction
In 2022, furrow-irrigated rice accounted for 18% of rice hectares in Arkansas (Hardke et al. Reference Hardke2022). The implementation of furrow irrigation involves drill-seeding rice on raised beds similar to methods used in corn, soybean, and cotton production in the mid-southern United States (Chlapecka et al. Reference Chlapecka, Hardke, Roberts, Mann and Ablao2021; Norsworthy et al. Reference Norsworthy, Scott, Bangarwa, Griffith, Wilson and McClelland2011b). Unlike flood-irrigated rice, which is typically flooded after it reaches the V5 growth stage, furrow-irrigated rice involves administering water through the furrows via polyethylene pipe and using gravity to move it away from the higher end of the field (Bagavathiannan et al. Reference Bagavathiannan, Norsworthy and Scott2011; Counce et al. Reference Counce, Keisling and Mitchell2000). Although producing furrow-irrigated rice can be advantageous over a flooded rice system, grain yields in a flooded rice system generally exceed those of furrow-irrigated rice (Vories et al. Reference Vories, Counce and Keisling2002). With effectively managed furrow-irrigated rice, growers can decrease labor and input costs depending on soil texture, topography, and other climatic barriers by using up to 23% less water relative to a delayed-flood system (Chlapecka et al. Reference Chlapecka, Hardke, Roberts, Mann and Ablao2021; Massey et al. Reference Massey, Reba, Adivento-Borbe, Chiu and Payne2022).
The water management practices associated with different rice production systems can also significantly influence the weed spectrum present in a field (Kraehmer et al. Reference Kraehmer, Jabran, Mennan and Chauhan2016). In a delayed-flood system, terrestrial weed emergence typically occurs before flooding due to the anaerobic conditions acting as a weed suppression mechanism. Still, the intrinsic nature of furrow-irrigated rice enables weed emergence throughout most of the growing season due to a consistently wet environment (Bagavathiannan et al. Reference Bagavathiannan, Norsworthy and Scott2011). In flooded rice, barnyardgrass [Echinochloa crus-galli (L.) P. Beauv.], sedges (Cyperus spp.), and weedy rice (Oyrza sativa L.) are among the most problematic weed species (Butts et al. Reference Butts, Kouame, Norsworthy and Barber2022). However, furrow-irrigated rice creates a favorable environment for traditional upland crop weeds, such as Palmer amaranth, to flourish throughout the growing season (Beesinger et al. Reference Beesinger, Norsworthy, Butts and Roberts2022; Norsworthy et al. Reference Norsworthy, Scott, Bangarwa, Griffith, Wilson and McClelland2011b).
Rice, like many other agronomic crops, can be a host for numerous broadleaf and grass species. In 2020, survey respondents indicated that Palmer amaranth was the second and fifth most troublesome weed species in furrow- and flood-irrigated rice, respectively, with barnyardgrass holding the top position in both systems (Butts et al. Reference Butts, Kouame, Norsworthy and Barber2022). The increased adoption of furrow irrigation enhances potential problems with Palmer amaranth due to the extended emergence period the system provides (Norsworthy et al. Reference Norsworthy, Griffith and Scott2008). While information is scarce on the effect of Palmer amaranth on rice yields, the competitive nature of the weed has been reported in cotton and soybean production in the mid-southern United States because it ranks among the most troublesome weeds in both crops (Klingaman and Oliver Reference Klingaman and Oliver1994; Van Wychen Reference Van Wychen2022). With an increasing number of hectares being used for furrow-irrigated rice and the innate combative character of Palmer amaranth, a dire demand exists for methods to control the weed in rice production.
Herbicides typically serve as the foundation of a weed control program due to their ease of application and general effectiveness against problematic weeds (Norsworthy et al. Reference Norsworthy, Ward, Shaw, Llewellyn, Nichols, Webster, Bradley, Frisvold, Powles, Burgos, Witt and Barret2012; Priess et al. Reference Priess, Norsworthy, Godara, Mauromoustakos, Butts, Roberts and Barber2022). Unfortunately, Palmer amaranth has evolved resistance to many herbicide sites of action that are typically applied to rice, meaning a creative weed management program that includes multiple control methods is vital for successful weed control (Norsworthy et al. Reference Norsworthy, Griffith and Scott2008, Reference Norsworthy, Korres, Walsh and Powles2016). Chemical, cultural, biological, and physical control practices are key factors in an integrated weed management program, which supports the suggested zero-tolerance threshold associated with Palmer amaranth management (Norsworthy et al. Reference Norsworthy, Griffith, Griffin, Bagavathiannan and Gbur2014; UC IPM 2020). Broadening weed control practices is important because repeated herbicide use poses the potential to become less effective due to the increasing incidences of herbicide-resistant Palmer amaranth (Norsworthy et al. Reference Norsworthy, Ward, Shaw, Llewellyn, Nichols, Webster, Bradley, Frisvold, Powles, Burgos, Witt and Barret2012). Additionally, even under favorable conditions and timely applications, herbicides rarely provide complete weed control (Bagavathiannan and Norsworthy Reference Bagavathiannan and Norsworthy2012). Weed control options in furrow-irrigated rice, outside the scope of herbicide chemistries, need to be identified to reduce the risk of herbicide resistance in troublesome weed species.
One way to diversify a weed management program is through the use of cover crops. Implementing cultural control methods such as planting cover crops helps to minimize reliance on herbicides and shift the focal point to reducing weed emergence from the soil seedbank (Shekhawat et al. Reference Shekhawat, Rathore and Chauhan2020). Winter-annual cover crop usage was initially targeted for improving soil health and preventing surface runoff; however, potential weed control benefits from cover crops have been demonstrated in recent years (Krutz et al. Reference Krutz, Locke and Steinriede2009; Norsworthy et al. Reference Norsworthy, McClelland, Bangarwa and Still2011a; Price et al. Reference Price, Balkcom, Duzy and Kelton2012). In cotton and soybean production systems, cover crops can assist in reducing Palmer amaranth emergence (DeVore et al. Reference DeVore, Norsworthy and Brye2012; Palhano et al. Reference Palhano, Norsworthy and Barber2018).
In Arkansas, cereal rye and winter wheat can reduce Palmer amaranth emergence by up to 83% and 78%, respectively, compared to treatments that omit the use of cover crops (Palhano et al. Reference Palhano, Norsworthy and Barber2018). The chemical and physical characteristics of cover crop residues reduce weed seed germination (Liebl et al. Reference Liebl, Simmons, Wax and Stoller1992; Moore et al. Reference Moore, Gillespie and Swanton1994). Furthermore, some types of cereal rye can lower Palmer amaranth germination and development through the innate ability to produce allelopathic chemicals such as 2,4-dihydroxy-1,4(2H)-benzoxazine-3-1 and 2,3-benzoxazolinone, during residue decomposition (Burgos and Talbert Reference Burgos and Talbert2000; Webster et al. Reference Webster, Scully and Culpepper2013). Legume cover crops such as crimson clover (Trifolium incarnatum L.) and hairy vetch can also reduce weed emergence through the production of allelopathic compounds (Fisk et al. Reference Fisk, Hesterman, Shrestha, Kells, Harwood, Squire and Shaeffer2001; White et al. Reference White, Worsham and Blum1989). While cover crop usage in rice production is novel, the cultural practice could prove beneficial in providing early-season suppression of Palmer amaranth, potentially eliminating preemergence herbicide applications and reducing input costs. Therefore, the objective of this study was to determine the best cover crop for suppression of Palmer amaranth while having the least effect on rice.
Materials and Methods
Influence of Cover Crops on Palmer Amaranth Suppression and Rice Development
Field experiments were initiated at the Lon Mann Cotton Research Station in Marianna, AR (34.72567°N, 90.73498°W), in 2022, and the Milo J. Shult Research and Extension Center in Fayetteville, AR (36.09344°N, 94.17449°W), in 2022 and 2023. One trial for a given experiment focused on cover crop biomass, rice stand establishment, rice groundcover, and rough rice grain yield assessments while being kept free of all weeds. An identical, adjacent trial with the same experimental setup focused on the suppression of Palmer amaranth emergence. In the fall, before each rice growing season, the ground was tilled and hipped into 96-cm spaced and 91-cm spaced beds in Marianna and Fayetteville, respectively. The soil at the Marianna location was a Convent silt loam (course-silty, mixed, superactive, nonacid, thermic Fluvaquentic Endoaquepts) consisting of 9% sand, 80% silt, 11% clay, and 1.8% organic matter with a pH of 6.5. In Fayetteville, the soil was a Leaf silt loam (fine, mixed, active, thermic Typic Albaquults) consisting of 18% sand, 69% silt, 13% clay, and 1.6% organic matter, with a pH of 6.6. The experiments were conducted as a randomized complete block design with five monoculture cover crop treatments, each replicated four times.
In the fall, plots were drill-seeded with cover crops on a 19-cm spacing, which included cereal rye, wheat, Austrian winterpea, and hairy vetch. A control plot with no cover crop planted was included for comparison. Cereal rye, wheat, Austrian winterpea, and hairy vetch were sown at 67, 67, 50, and 17 kg ha−1, respectively (Roberts Reference Roberts2021). At all rice plantings, a hybrid, long-grain rice cultivar ‘RT 7321FP” (RiceTec Inc., Alvin, TX) was planted at 36 seeds m−1 of row at a 1-cm depth with 19 cm between rows. In 2022, rice was planted in separate experiments on April 22 and May 3 in Fayetteville, and an additional site in Marianna on April 27. In 2023, rice was planted on April 15 and May 3 in separate experiments in Fayetteville. In total, this experiment consisted of five site-years (Table 1). Plot dimensions were 3.9 m wide (four beds) by 5.2 m long in Marianna, and 3.7 m wide (four beds) by 5.2 m long in Fayetteville. A 0.9-m alley was placed between blocks. All herbicides, including over-sprays, were applied using a CO2-pressurized backpack sprayer calibrated to deliver 140 L ha−1 at 276 kPa using four AIXR 110015 nozzles (TeeJet Technologies, Glendale Heights, IL) at 4.8 km h−1. The soil for each trial was amended for fertility before planting based on soil test values provided by the Marianna Soil Test Laboratory. Nitrogen, as urea (460 g N kg−1), was applied at 135 kg N ha−1 in three separate applications at 2-wk intervals beginning at the V5 stage of rice.
Table 1. List of dates for cover crop planting and termination and rice planting for each site-year. a
a Calendar year that the cover crop was terminated and rice was planted.
Two weeks before each rice planting and at planting, each trial received an application of glyphosate at 1,260 g ae ha−1 for cover crop termination. Clomazone at 336 g ai ha−1 was broadcast-applied to all experiments on the day of rice planting for residual control of annual grasses. Aboveground cover crop biomass was collected from two 0.5-m2 quadrats within the center two rows of each plot before planting rice. All harvested aboveground biomass was placed in an oven at 66 C for 2 wk, dried to constant mass, and then weighed. Seven days after rice emergence, rice plants in two 1-m sections of row were counted in each plot. Singular images (in red-green-blue) of the experiment were captured at 40 m above the crop canopy by an unmanned aerial system (DJI Mavic 2; DJI Technology Co., Nanshan, Shenzhen, China) 2 wk before heading and used to determine rice groundcover based on green pixel counts using Field Analyzer (Green Research Services, Fayetteville, AR).
An earlier greenhouse experiment confirmed that propanil and thiobencarb do not affect Palmer amaranth emergence (personal observations). Hence, propanil (STAM; UPL, King of Prussia, PA) was applied approximately three times during the 5 wk following rice emergence in the experiment in which Palmer amaranth density was quantified. After rice planting, two 1-m2 quadrats were established in each plot, allowing Palmer amaranth plants to be counted weekly and removed 5 wk after rice emergence. The yield assessment portion of the experiment relied on applications of fenoxaprop-p-ethyl (Ricestar® HT; Gowan Co., Yuma, AZ) and florpyrauxifen-benzyl (Loyant®; Corteva Agriscience, Indianapolis, IN), as well as hand-weeding, to keep the experimental area weed-free throughout the growing season. After the rice reached maturity, the center two rows of each four-row plot were harvested using an 8-XP plot combine (Kincaid, Haven, KS) with a header width of 1.8 m. The yields collected from each plot were adjusted to 12% moisture.
Data Analysis
Data were analyzed using JMP Pro software (v. 17.0; SAS Institute Inc, Cary, NC). Data were assessed for normality using Shapiro-Wilk tests and equal variance by testing the residuals in the distribution platform with JMP software. Data that did not satisfy normal distribution and equal variance assumptions were analyzed using the GLIMMIX procedure in JMP. After the residuals failed to violate the Shapiro-Wilk and Levene tests, cover crop biomass, rice stand, ground coverage, and yield were analyzed using a Gaussian or normal distribution, whereas Palmer amaranth count data assumed a Poisson distribution (Gbur et al. Reference Gbur, Stroup, McCarter, Durham, Young, Christman, West and Kramer2012). All data were subjected to ANOVA to evaluate the main effect of cover crop, and means were separated using Fisher’s protected LSD at value of α = 0.05. Normally distributed data were analyzed within JMP Pro using the fit-model platform, and Palmer amaranth count data were analyzed using the generalized linear mixed model add-in with a Poisson distribution (Gbur et al. Reference Gbur, Stroup, McCarter, Durham, Young, Christman, West and Kramer2012). Site was analyzed separately due to differences in weed density between the experimental locations.
Results and Discussion
Cover Crop Biomass
Cover crop biomass differed by cover crop treatment in two of the five site-years, and dry biomass ranged from 430 to 3,440 kg ha−1 across all studies (Table 2), which is similar to the range of biomass that Wiggins et al. (Reference Wiggins, Hayes, Nichols and Steckel2017) observed with similar cover crop treatments. In 2022, cereal rye produced 2,680 kg ha−1 of biomass, which was greater than that of Austrian winterpea and hairy vetch at 430 and 1,310 kg ha−1, respectively, at Marianna. Similarly, in 2023, biomass accumulation of Austrian winterpea (1,080 kg ha−1) and hairy vetch (1,070 kg ha−1) was less than that of cereal rye (1,560 kg ha−1) at Fayetteville of rice planted in April. Previous literature has shown that cereal cover crops, including cereal rye and winter wheat, produce greater aboveground biomass relative to legume cover crops such as hairy vetch, Austrian winterpea, and crimson clover (Daniel et al. 2019; Schulz et al. Reference Schulz, Marocco, Tabaglio, Macias and Molinillo2013).
Table 2. Influence of cover crop within five site-years on cover crop aboveground biomass, rice density, relative rice groundcover, relative grain yield, and Palmer amaranth density. a
a Means within a column and assessment followed by the same letter are not different according to Fisher’s protected LSD (α = 0.05).
Cereal rye produced more biomass than any other cover crop at the Marianna location in 2022, and in all other site-years, no cover crop produced more biomass than cereal rye (Table 2), creating a favorable environment for suppression of Palmer amaranth (Norsworthy et al. Reference Norsworthy, McClelland, Bangarwa and Still2011a). These results are similar to findings reported by Wiggins et al. (Reference Wiggins, Hayes, Nichols and Steckel2017), who also determined that cereal rye provided the greatest quantity of biomass. However, the extreme variability associated with cover crop biomass accumulation has been documented over several years and soil textures, indicating that the weed control efficacy is potentially less consistent in the absence of herbicides (Norsworthy et al. Reference Norsworthy, McClelland, Bangarwa and Still2011a; Palhano et al. Reference Palhano, Norsworthy and Barber2018). In three out of the five trials conducted, no differences in biomass production were observed among cover crops. In general, cover crop biomass was highly variable among trials, which could be attributed to the timing of cover crop planting, differences in heat unit accumulation in the spring, and cumulative precipitation (Grint et al. Reference Grint, Arneson, Oliveira, Smith and Werle2022; Mirsky et al. Reference Mirsky, Curran, Mortensen, Ryan and Shumway2011; Wilson et al. Reference Wilson, Baker and Allan2013). As a result, cover crop biomass production and subsequent efficacy in this region should be further evaluated, considering the growth and development can be dependent upon location (Schomberg et al. Reference Schomberg, McDaniel, Mallard, Endale, Fisher and Cabrera2006).
Rice Density
In four of the five site-years for the experiment, the main effect of cover crop did not influence rice establishment 7 d after emergence relative to the no cover crop treatment (P > 0.05) (Table 2), indicating that rice emergence is generally uninterrupted by the cover crops that were evaluated. Although in one trial, rice stand was reduced by 25%, 22%, 30%, and 22% relative to the no-cover-crop treatment with cereal rye, wheat, Austrian winterpea, and hairy vetch, respectively. Previous literature has documented that cover crops with high biomass have the potential to negatively affect planting, subsequently affecting crop uniformity and crop development (Kornecki et al. Reference Kornecki, Price, Raper and Arriaga2009; Schulz et al. Reference Schulz, Marocco, Tabaglio, Macias and Molinillo2013). In general, the cover crops evaluated in this study had minimal effect on rice emergence and establishment.
Relative Rice Groundcover
The main effect of cover crop did not influence rice groundcover in four of five trials for the experiment, with rice groundcover ranging from 93% to 102% at those sites relative to the no-cover-crop treatment (P > 0.05) (Table 2). However, rice groundcover was reduced following cereal rye by up to 13% relative to the no-cover-crop treatment in Fayetteville in 2023, when rice was planted in April. Wheat, with similar biomass production, did not reduce rice groundcover within the trial, suggesting other factors could have influenced the lack of soil coverage observed in the cereal rye treatments. The ability of cereal rye to release allelochemicals such as benzoxazinones is known to reduce crop growth and development; therefore, rice groundcover could be influenced by the production of these phytotoxic compounds (Martinez-Feria et al. Reference Martinez-Feria, Dietzel, Liebman, Helmers and Archontoulis2016). Additionally, controlled experiments in a laboratory once determined that allelochemicals were more harmful to small-seeded plant species (Liebman and Sunberg Reference Liebman and Sunberg2006). Cereal rye can also efficiently sequester nutrients within the soil, potentially affecting crop maturity through increased competition for soil minerals (Krueger et al. Reference Krueger, Ochsner, Porter and Baker2011). However, in most instances, rice development and canopy closure were unaffected by cover crop biomass production.
Relative Rice Grain Yield
The biomass produced by cover crops did not influence rough rice grain yield across all trials for the experiment (P > 0.05) (Table 2). One of the benefits of using a legume cover crop is the ability to fixate atmospheric nitrogen during plant decomposition, which then becomes available to the crop (Reddy Reference Reddy2001). To our knowledge, no peer-reviewed data have been published on the influence of cover crops on furrow-irrigated rice yields; however, a preliminary study produced by Henry and Clark (Reference Henry and Clark2023) showed no statistical yield differences with the use of a cover crop blend consisting of annual rye (Lolium perenne L.), cowpea (Vigna unguiculata L.), crimson clover, and Daikon radish (Raphanus sativus L.) compared to a no-cover-crop treatment in a furrow-irrigated rice system. In all cases, rice maturity was not disrupted by the established cover crops.
Palmer Amaranth Density
Across all trials of the experiment, Palmer amaranth emergence varied by cover crop treatment (Table 2). On average, Palmer amaranth emergence was greater in trials conducted in 2022 (197 m−2) than in 2023 (32 m−2). In 2022, at the Marianna site, cereal rye and Austrian winterpea were the only cover crops to significantly reduce Palmer amaranth emergence, minimizing total emergence by 19% and 31% compared with the nontreated control, respectively. Oppositely, Palmer amaranth densities were comparable or greater among each cover crop treatment compared with those of the no-cover-crop treatment, with total weed emergence being greatest in both legume cover crop treatments at Fayetteville in 2022. Wiggins et al. (Reference Wiggins, Hayes and Steckel2016) also found that Palmer amaranth densities were comparable in several evaluated monoculture cover crops, including cereal rye, crimson clover, hairy vetch, and winter wheat. Additionally, Norsworthy et al. (Reference Norsworthy, McClelland, Giffith, Bangarwa and Still2010) reported that hairy vetch and Austrian winterpea provided minimal benefit in suppressing Palmer amaranth in cotton production in the mid-southern United States.
In 2023, at the April rice planting, Palmer amaranth emergence was not reduced by any of the cover crops compared with the no-cover-crop treatment. However, weed pressure at this location was extremely low compared to other sites, considering the nontreated control totaled 2.5 plants m2, on average, by the end of the evaluation period. Within the same year, at the May rice planting, a 35%, 82%, and 41% reduction in Palmer amaranth emergence was provided by cereal rye, Austrian winterpea, and hairy vetch, respectively. Considering soil disturbance can influence weed germination and emergence (Chauhan et al. Reference Chauhan, Gill and Preston2006), Palmer amaranth emergence in furrow-irrigated rice will likely be enhanced due to increased soil and cover crop residue disturbance from the narrow and more frequent row spacing in contrast to the typical planting methods used in corn, cotton, and soybean production.
Palmer amaranth densities were generally higher for cereal cover crops than legume cover crops. Legume cover crops innately possess lower carbon-to-nitrogen (C:N) ratios than cereal cover crops, allowing for less persistence on the soil surface; hence, Palmer amaranth suppression could be influenced by the increased decomposition rate of cover crop residues (Berg and McClaugherty Reference Berg and McClaugherty2003; Clark Reference Clark2012; Pittman et al. Reference Pittman, Barney and Flessner2020; Touchton et al. Reference Touchton, Rickerl, Walker and Snipes1984). Likewise, research has revealed that some Amaranthus species are extremely responsive to soil inorganic nitrogen, consequently increasing the competitive ability of the weed with the crop (Blackshaw et al. Reference Blackshaw, Brandt, Janzen, Entz, Grant and Derksen2003; Blackshaw and Brandt Reference Blackshaw and Brandt2008). Based on the Palmer amaranth data collected, cereal cover crops provide little to no benefit from a weed control standpoint in furrow-irrigated rice.
Practical Implications
Cover crops have proven to be an effective weed control tactic when targeting problematic weeds, including Palmer amaranth, in many cropping systems across the United States (Brennan and Smith Reference Brennen and Smith2005; Burgos and Talbert Reference Burgos and Talbert2000; Collins et al. Reference Collins, Chase, Stall and Hutchinson2007; Fisk et al. Reference Fisk, Hesterman, Shrestha, Kells, Harwood, Squire and Shaeffer2001; Palhano et al. Reference Palhano, Norsworthy and Barber2018; Reddy Reference Reddy2001). Only one research series publication has evaluated the effect of cover crops in a furrow-irrigated rice system, which constitutes approximately 18% of Arkansas rice hectares (Hardke et al. Reference Hardke2022; Henry and Clark Reference Henry and Clark2023). Although results differed among trials, the experiments conducted in 2022 and 2023 show some potential for cover crops to be used in furrow-irrigated rice to manage Palmer amaranth.
In most cases but not all, the cover crops evaluated in this study did not reduce rice emergence, groundcover, or grain yield. However, high biomass production from cover crops can affect crop emergence, as demonstrated here (Table 2) and in research conducted by Schulz et al. (Reference Schulz, Marocco, Tabaglio, Macias and Molinillo2013). Cereal rye appears to exhibit some ability to reduce Palmer amaranth emergence through increased biomass production, suppressing the weed by 19% to 35% in three of five trials. Additionally, legume cover crops generally decreased total Palmer amaranth emergence, with weed emergence being lowest for legume cover crops in three of the five site-years for the experiment.
Cover crop biomass accumulation and Palmer amaranth suppression from cover crops varied by site-year, suggesting that more research is needed in the rice-growing regions to ensure greater confidence before adopting as a stand-alone weed control method in furrow-irrigated rice production. Additionally, bed width is a key component in both weed and crop development due to its potential effects on irrigation and soil moisture content (Reed et al. Reference Reed, Butts, Norsworthy, Hardke, Barber, Bateman, Poncet and Kouame2024); hence, future cover crop research may also include determining the optimal bed width for increased Palmer amaranth suppression. Furthermore, some cover crop studies found added weed control when combining preemergence and postemergence herbicides with cover crops (Reddy Reference Reddy2001; Reeves et al. Reference Reeves, Price and Patterson2005; Wiggins et al. Reference Wiggins, McClure, Hayes and Steckel2015). Thus, it may be advantageous to replicate these experiments in conjunction with a standard rice herbicide program to determine whether the addition of herbicides would result in enhanced Palmer amaranth suppression compared to the observations noted here.
Funding statement
Partial support for this research was provided by the Arkansas Rice Research and Promotion Board.
Competing interests
The authors declare they have no competing interests.
Open access
